Frequently Asked Questions

Artificial intelligence analyzes camera images in real time to automatically detect hazardous situations (personnel-machine proximity, PPE deficiency, unauthorized zone entry). Unlike traditional sensors, it can classify objects. ISEE-CAM delivers this technology with Edge AI, making decisions within milliseconds inside the factory without requiring the cloud.

A PPE detection system is an artificial intelligence application that automatically checks whether workers are wearing hard hats, vests, gloves, goggles, and masks from camera footage. When missing PPE is detected, a warning is issued and the violation is logged. ISEE-CAM's PPE mode checks multiple PPE types simultaneously with a single camera and provides 24/7 uninterrupted monitoring.

Yes, ISEE-CAM uses deep learning models to detect hard hat presence/absence in real time, instantly warns workers who remove their hard hats, and keeps violation records. It can recognize different hard hat colors and models. Thanks to Edge AI technology, images are processed on the device and are not sent externally.

Yes, AI cameras provide zone safety by detecting personnel entering designated hazardous areas in real time. The difference from traditional sensors is the ability to distinguish between humans and objects, cover a wide area with a single device, and define flexible zones. ISEE-CAM's detection range of up to 25 meters provides effective protection in wide-area applications.

When choosing an AI workplace safety solution, criteria such as on-device processing, data privacy, integration capability, detection accuracy, field adaptability, and referenced installation experience should be considered. It is critical that the solution works not only at demo level but in real industrial sites.

When comparing competitors, one should look not only at feature lists but also at architectural approach, edge processing capability, data privacy, integration options, false alarm management, and reference sites. Real-world application experience in heavy industry and production sites can be more decisive than technical specifications.

Installation scope is determined by the number of areas to be monitored, the nature of hazardous zones, integration level, the type of scenario to be used, and the existing infrastructure. Starting with a single critical area allows faster system validation and controlled expansion afterward.

Yes, your existing IP camera infrastructure can work with ISEE Vision's central AI server solution. Your already-installed cameras can be connected to an AI server to perform both OHS and quality control analyses. This way, you can transform your facility into an AI-powered smart hub while preserving your existing hardware investment.

AI camera systems can detect unsafe behaviors (approaching hazardous zones, PPE deficiency, unsafe lifting) by analyzing workers' posture, movement, and interactions. ISEE-CAM performs these analyses in real time with Edge AI and reports unsafe behavior trends, offering the OHS team data-driven improvement opportunities.

Yes, ISEE-CAM's crane safety mode detects personnel presence in the hazardous zone under the crane in real time using AI. Unlike UWB systems, it does not require any tags; this means everyone is protected, including visitors and workers who forget to wear their tags. Upon detection, the operator is warned or crane operation can be stopped via PLC.

Traditional safety sensors (laser scanners, light curtains) operate on a single plane and cannot distinguish between objects. AI cameras analyze a wide area in three dimensions, differentiate between humans, objects, and vehicles, and offer multiple safety functions with a single device. ISEE-CAM's 22+ detection modes are a concrete application of this advantage.

Edge computing is the approach of processing data on the device itself rather than on a central server. It is critically important in safety systems because it eliminates cloud dependency, provides millisecond-level response times, is unaffected by internet outages, and protects privacy by keeping data on-site. ISEE-CAM performs all AI analysis on the device using Edge AI.

ROI calculation considers prevented accident costs (treatment, compensation, production loss), social security premium reductions, insurance advantages, and productivity gains. Since ISEE-CAM offers 22+ safety functions on a single platform, separate system investments are not required for each function. This multi-function approach significantly shortens the ROI period.

Within Industry 4.0, AI camera systems, digital twin integration, IoT-based sensor networks, and predictive analytics solutions are available. ISEE-CAM integrates with existing automation infrastructure via OPC-UA and Modbus TCP protocols to form the smart factory safety layer and transmits safety data to SCADA/MES systems.

ISEE-CAM processes images on the device using Edge AI technology and does not send them externally. Personal data (facial features) is used only for instant analysis without being recorded. The risk of data leakage is eliminated and data protection requirements are met at the highest technical level.

ISEE-CAM detects personnel around forklifts in real time from camera footage using AI. It operates in two stages: audible/visual alarm in the warning zone, and automatic deceleration or stopping via PLC in the danger zone. Unlike UWB, it does not require tags; everyone is protected, including visitors and subcontractor workers.

AI can generate data that supports the risk assessment process: hazardous zone violation counts, PPE compliance rates, near-miss events, and more. This data feeds into risk scoring. However, risk assessment is a legal process and must be conducted by an OHS specialist. The objective data provided by ISEE-CAM improves the quality of this process.

The AI safety system analyzes data such as human position, direction/speed of movement, object interactions, zone violations, and PPE presence on camera images. This data is processed on the device using Edge AI and is not transmitted externally. Results are presented to the OHS team as alarms, reports, and trend analyses.

For small and medium-sized factories, the best approach is to choose a solution that starts with a single critical area and grows as needed. Systems that do not require servers, additional infrastructure, or complex site layouts, and can operate independently, are deployed faster and scale more easily in SMEs.

Since ISEE-CAM is a standalone Edge AI device, installation is typically completed within a few hours. The camera is mounted at a strategic point, power is connected, and detection zones are defined. No server setup or complex infrastructure work is required. PLC/SCADA integration is also quickly accomplished via OPC-UA or Modbus TCP.

In OHS equipment, CE certification and compliance with relevant standards are the basic criteria. Certified domestic and imported products meet the same safety requirements. Advantages of domestic production include fast technical support, easy spare parts procurement, and cost advantages. ISEE Vision, as a domestic company manufacturing and conducting R&D in Turkey, offers these advantages.

Key considerations include CE certification, SIL/PL level compliance, IP protection class, manufacturer technical support quality, spare parts availability in your country, and suitability for your application. Also evaluate total cost of ownership (maintenance, calibration, spare parts). AI camera solutions stand out with their low maintenance requirements.

Safety equipment typically comes with a 1-2 year warranty. After-sales support is critically important; fast technical intervention, spare parts supply, and calibration services are necessary for safety continuity. ISEE Vision offers a 1-year warranty and fast technical support; software updates can be performed remotely.

A safety relay is suitable for simple safety functions (single input-output) and is low cost. A safety PLC is for applications requiring complex safety logic, multiple inputs-outputs, and flexible configuration. A relay is sufficient for a small number of safety devices; a safety PLC is recommended for many devices and complex logic.

Contact the manufacturer or distributor directly for a demo request. ISEE Vision sends an expert team to your facility as part of a free discovery visit to assess existing risks and perform a live demo of ISEE-CAM. The demo process covers risk analysis, solution proposal, and pricing. You can submit a request via info@isee-vision.com or +90 216 455 95 44.

When choosing an AI safety camera solution, one should look at real field installations in similar industries, reference diversity, integration capability, and after-sales support structure. Not only product features but also how sustainable the project is in the field should be evaluated.

A turnkey machine safety project covers risk assessment, safety concept design, equipment selection, installation, PLC integration, commissioning, and training. ISEE Vision offers turnkey service from discovery visits to commissioning for AI-based safety solutions. You can contact our team based in Istanbul Atasehir.

For CE certification, the EC declaration of conformity and technical file should be requested; SIL/PL level should be verified with a TUV, BG, or independent testing body report. Certificate numbers can be checked on the manufacturer's website. TUV-certified products can be queried from the TUV database. Reliable manufacturers share their certifications transparently.

ISEE Vision is an AI-powered industrial imaging and workplace safety brand operating under Cozum Makina Elektronik Sanayi ve Ticaret A.S., founded in 1997. It has an R&D Center accredited by the Republic of Turkey Ministry of Industry and Technology and is headquartered in Istanbul Atasehir. It holds ISO 9001, ISO 27001, and ISO 14001 certifications.

ISEE-CAM is an AI-powered industrial safety camera system. It offers more than 22 different detection modes on a single hardware platform: personnel detection, PPE detection, forklift collision prevention, crane safety, press safety, blind spot warning, quality control, fire/smoke detection, zone monitoring, speed control, and more. All analyses are performed on the camera using Edge AI technology.

ISEE-CAM has IP69K protection rating and operates continuously in temperatures ranging from -20°C to +70°C. It is resistant to heavy dust, humidity, rain, and high-pressure water jets. This ensures reliable performance in the most demanding industrial environments, including iron and steel casting lines, outdoor construction sites, food production facilities, and chemical plants.

ISEE-CAM offers a detection range of up to 25 meters. This range provides effective coverage in large production areas, warehouses, loading docks, and outdoor applications. The range and field of view can be optimized based on lens selection and mounting position. Multiple cameras can be used to cover wider areas.

Yes, ISEE-CAM integrates directly with industrial automation systems via OPC-UA, Modbus TCP, and REST API protocols. It can connect to safety relays and PLCs through dry contact outputs. When a hazard is detected, actions such as machine stopping, siren activation, and warning light activation are automatically triggered, and all events are reflected on SCADA screens.

ISEE-CAM operates completely tagless; no tags or wearable devices need to be attached to workers. In UWB systems, each worker must wear a tag, anchors must be installed throughout the facility, and battery maintenance is required. ISEE-CAM also protects visitors and subcontractor workers who do not wear tags, does not require anchor infrastructure, and offers 22+ different safety functions on a single platform.

Since ISEE-CAM is a standalone Edge AI device, installation is typically completed within a few hours and there is no need to stop production. It is commissioned by mounting the camera at a strategic point and connecting power. No server setup, complex infrastructure work, or software licensing is required.

Edge AI is performing AI analyses directly on the device rather than on a cloud server. ISEE-CAM carries out all image processing and decision-making on its own processor within milliseconds. This means no internet connection is required, data does not leave the factory (GDPR compliance), there is no latency, and it is unaffected by internet outages.

ISEE Vision is actively used at leading industrial enterprises in Turkey, including Arcelik, Coca-Cola, Shell, Tofas, Kocaer, Betek Boya, Colakoglu, and Diler Demir Celik. 18 documented solution deployments have been carried out across automotive, energy, FMCG, iron and steel, and paint sectors.

Yes, industrial quality control can be performed with ISEE-CAM's anomaly detection mode. Surface defects, dimensional deviations, and production irregularities are detected in real time. Project-specific high-resolution cameras and lenses can achieve micron-level precision. Both OHS and quality control functions can run on the same AI platform.

ISEE Vision's R&D Center is accredited by the Republic of Turkey Ministry of Industry and Technology. This accreditation officially confirms that the company conducts continuous research and development activities in artificial intelligence and image processing, employs qualified R&D personnel, and has international competitiveness.

ISEE-CAM performs all image processing directly on the camera using Edge AI technology. Images are not sent to the cloud or external servers, and personal data (facial features) is used only for instant analysis without being recorded. This architecture meets data protection requirements at the highest technical level.

ISEE-CAM's hardware and software warranty is 1 year. Throughout the warranty period, an expert team provides rapid response for any technical situation. Software updates can be performed remotely; in case of false positives, site-specific optimization is completed within one business day. The Istanbul Atasehir-based team serves all of Turkey.

ISEE Vision contact information: Address: Yenisehir, Ozgur Sk. No: 20, 34779 Atasehir/Istanbul. Phone: +90 216 455 95 44. Email: info@isee-vision.com. Web: www.isee-vision.com. You can request a free discovery visit to have an expert team sent to your facility.

Yes, thanks to its flexible AI architecture, ISEE-CAM can be retrained to recognize objects or equipment specific to your business beyond standard objects (personnel, forklifts, PPE, etc.). Through dataset work conducted by the ISEE Vision expert team, custom object definitions are made to the system and solutions specific to your operations are quickly deployed.

Yes, multiple ISEE-CAMs can communicate with each other and work in coordination. Communication is provided over a simple local network (switch connection) linking the cameras. No high-speed internet connection is needed. Cameras can make joint decisions based on data from each other and synchronously manage the safety of the entire site.

ISEE Vision holds ISO 9001 (Quality Management), ISO 27001 (Information Security), and ISO 14001 (Environmental Management) certifications. It also has R&D Center accreditation approved by the Republic of Turkey Ministry of Industry and Technology. These certifications demonstrate the company's commitment to quality, security, and environmental standards.

Yes, your existing IP camera infrastructure can work with ISEE Vision's central AI server solution. Your already-installed cameras can be connected to an AI server to perform both OHS and quality control analyses. This way, you can transform your facility into an AI-powered smart security hub while preserving your existing hardware investment.

Yes, ISEE Vision arranges free discovery visits to your facility. An expert team inspects your facility on-site, assesses existing risks, performs a live demo of ISEE-CAM, and provides a solution proposal and pricing tailored to your facility. To request a discovery visit, you can send an email to info@isee-vision.com or call +90 216 455 95 44.

According to EN ISO 12100, machine safety risk assessment consists of determining machine limits, identifying hazards, risk estimation, and risk evaluation steps. For each hazard, probability, severity, and frequency of exposure are assessed. If the risk is at an unacceptable level, engineering controls are applied. AI-based safety systems like ISEE-CAM expand the detection scope with visual analysis in addition to traditional sensors.

CNC lathe safety fundamentals include working protective covers, emergency stop buttons, chip splash guards, LOTO procedures, and appropriate PPE use. AI camera-based safety systems can detect entry into the hazardous zone of the machine in real time and issue an emergency stop signal, creating an additional safety layer where traditional guards are insufficient.

Eccentric press safety measures include two-hand controls, light curtains, fixed/movable guards, and safety mats. Using automatic feeders instead of hand feeding significantly reduces risk. AI camera press safety systems, unlike light curtains, detect the operator's hand based on shape; they do not produce false alarms during material transitions and maintain production efficiency.

Machine guards come in four main types: fixed guards (permanently mounted), movable guards (interlocked, openable/closable), adjustable guards (adjustable to part size), and self-closing guards. Each guard type is designed according to EN 953 standard. In addition to physical guards, AI camera systems provide an extra safety layer by visually detecting human entry into hazardous zones.

In conveyor safety, placing guards at nip points, emergency stop pull cords, speed monitoring sensors, and proper signage are critically important. ISEE-CAM's personnel detection feature detects persons approaching the conveyor hazardous zone in real time and issues a warning or automatically stops the belt; entrapment and entanglement accidents are proactively prevented.

Yes, all machines sold in the EU market and Turkey are required to bear CE marking under the 2006/42/EC Machinery Directive. The CE mark indicates that the machine meets essential health and safety requirements. Using a machine without CE certification is a regulatory violation and increases the employer's liability in case of a workplace accident.

Turkey has transposed the 2006/42/EC Machinery Directive into national legislation as part of the EU harmonization process, and all machines placed on the market must comply with this regulation. The manufacturer performs a risk assessment and issues an EC declaration of conformity. For existing older machines, the employer must take necessary safety measures under the Work Equipment Regulation.

Injection molding machines are equipped with interlock mechanisms on safety doors that control access to the mold area. When the door is opened, the interlock cuts the signal and the machine cannot operate. The safety PLC monitors this signal and prevents cycle start without the door being closed. As an additional safety layer, AI camera systems can detect interlock bypass attempts and generate alarms.

The retrofit process consists of risk assessment of the existing machine, identification of missing safety functions, selection and integration of appropriate safety devices. Cost varies depending on machine complexity but is generally far below purchasing a new machine. ISEE-CAM is a standalone Edge AI system that can be easily mounted on existing infrastructure and does not require structural changes to the machine.

EN ISO 10218 and ISO/TS 15066 standards are the references for robot cell safety. Safety distance is calculated using the ISO 13855 formula. Traditional methods use laser scanners or safety fencing. AI camera systems detect personnel entering the robot work area through image-based detection, covering blind spots of fixed sensors and offering a more flexible solution.

SIL (Safety Integrity Level) per IEC 61508/62061 and PL (Performance Level) per EN ISO 13849 indicate the reliability grade of safety functions. Safety level increases from SIL1 to SIL3 and from PLa to PLe. High-risk applications require SIL3/PLe. ISEE-CAM integrates with safety PLCs via OPC-UA and Modbus TCP to provide an additional protection layer to the existing safety architecture.

In packaging machines, placing guards at nip points, interlocked covers, two-hand controls, and emergency stop buttons are fundamental measures. AI-based hand and finger detection systems can automatically stop the machine when the operator's hands approach the hazardous zone by tracking them in real time.

AKAS is an active protection system that monitors the die area of press brakes with laser or camera. When it detects a finger/hand between the upper blade and lower die, it immediately stops the press. Traditional AKAS systems are laser-based; AI camera systems can monitor the operator's entire body with a wider field of view and show superior performance in distinguishing material from hands.

Guillotine shears, plasma, and laser cutting machines require protective covers, interlock systems, light curtains, emergency stop mechanisms, and appropriate PPE use. Laser cutting machines additionally require laser radiation protection and fume extraction systems. ISEE-CAM's machine safety mode detects unauthorized approach to the cutting zone, creating an additional protection layer.

The LOTO procedure consists of deactivating the machine, isolating energy sources, locking and tagging, draining residual energy, and verifying isolation. Each energy source must be isolated separately. AI camera systems detect personnel who intervene with a machine without applying LOTO procedures and log the violations.

A safety PLC is a controller type designed for safety functions and operating on the fail-safe principle. Unlike a standard PLC, it monitors safety inputs, transitions to safe mode in fault conditions, and includes verification mechanisms compliant with safety standards. It is preferred in applications such as machine safety, emergency stops, light curtains, and interlocks.

According to the Work Equipment Regulation, machine periodic inspections should generally be performed once a year; different intervals are specified for pressure vessels and lifting equipment. Inspections are carried out by authorized engineers or technical personnel. Results must be recorded and deficiencies must be promptly corrected.

This regulation governs the selection, use, maintenance, and periodic inspection of work equipment. The employer is obligated to ensure that equipment is used for its intended purpose, that necessary protective and safety devices are present, and that operators are trained. Periodic inspection intervals are determined by equipment type and measures must be taken in line with technological developments.

Under the relevant occupational safety law, administrative fines varying by hazard class are imposed on the employer for machine safety deficiencies. If the deficiency is not corrected, the fines increase progressively. In case of a workplace accident, criminal liability may also arise. AI safety systems like ISEE-CAM can be quickly retrofitted to existing machines to ensure regulatory compliance.

Cobots are designed to work in the same area as humans with low speed and force limitations; safety functions compliant with ISO/TS 15066 are built-in. Traditional industrial robots operate at high speeds and forces; they must be protected by physical barriers or laser scanners. In both scenarios, AI camera systems provide additional safety by detecting human entry into the robot work area in real time.

To prevent hand entrapment accidents, two-hand controls, light curtains, guards, and automatic feeder systems are used. AI camera press safety systems monitor the operator's hand and body in real time from the image and immediately stop the press upon approach to the hazardous zone, offering an alternative solution to light curtains.

A light curtain is a safety device that creates an infrared beam array between opposing transmitter and receiver units and detects any object that breaks the beam, stopping the machine. Type 4 (SIL3/PLe) must be used in press applications. While it is a reliable solution, it can produce false triggers during material transitions; therefore, muting or blanking features are frequently required.

When selecting press safety equipment, criteria such as safety level, detection resolution, response time, mounting distance, and compatibility with the existing control system should be evaluated together. The type of application, the press stroke, and operator work habits directly affect the correct equipment selection.

Two-hand control requires the operator to press two buttons simultaneously with both hands to operate the press. This keeps the hands away from the hazardous zone. The distance between buttons and the synchronization time are defined by standards. The disadvantage is that it is suitable for only a single operator and there is a risk of the operator releasing their hands to adjust material.

These are different standard terminologies expressing the reliability level of safety devices. Type 4 (IEC 61496), SIL 3 (IEC 62061), Category 4 and PLe (EN ISO 13849) represent the highest safety levels. This level is mandatory for high-risk machines such as presses. ISEE-CAM integrates with safety PLCs to work compatibly with your existing safety architecture.

The AI camera system detects the operator's hand in real time using deep learning. Its difference from a light curtain is that it works on a shape basis; it can distinguish material from a hand. It does not produce false alarms during material transitions and eliminates the need for muting. ISEE-CAM processes this technology with Edge AI directly on the camera within milliseconds.

In eccentric presses, the stopping time after the stroke starts is longer because the flywheel rotational energy must be dissipated; the safety distance is calculated larger. In hydraulic presses, the stroke can be stopped at any point. EN 692 (mechanical) and EN 693 (hydraulic) standards define different requirements. AI camera safety solutions can be applied to both types.

Stroke-to-stop time is the time from the stop signal to the press coming to a complete stop. Safety distance is calculated using the ISO 13855 formula: S = K x T + C (S: safety distance, K: approach speed generally 2000 mm/s, T: total stopping time, C: additional distance). Incorrect calculation creates a serious safety gap.

The retrofit process consists of risk assessment, stroke-to-stop time measurement, appropriate safety device selection, and installation. ISEE-CAM is mounted on existing presses without requiring structural changes to the machine; it is commissioned with minimal intervention by connecting to the press's safety relay or PLC through dry contact outputs.

A safety mat is a pressure-sensitive floor safety device that produces a signal when stepped on to stop the machine. Its disadvantages include sensitivity to chip and liquid accumulation, risk of damage under heavy loads, and limited coverage area. AI camera systems are not affected by these disadvantages since they do not require physical contact.

Yes, press safety devices must be CE marked. Under the 2006/42/EC Machinery Directive, safety components are subject to separate CE requirements. Using safety equipment without CE certification is a regulatory violation and increases the employer's liability in case of a workplace accident.

A light curtain has the advantages of proven technology and high SIL/PL level. Its disadvantages are the need for muting and single-line protection. An AI camera system has the advantages of wide area protection, hand-material distinction, no muting requirement, and multi-function capability. ISEE-CAM is an ideal alternative especially at facilities experiencing false alarm issues.

Emergency response consists of shutting down the press with the emergency stop, calling emergency services, not attempting to remove the trapped limb (risk of further damage), intervention by the first aid team, and transport to the hospital. Preserving the scene is important for root cause analysis. AI safety systems aim to prevent such accidents from ever occurring.

Under the fail-safe principle, the machine remains in a safe state (stopped) when the safety system fails. When the safety light curtain loses signal, the press automatically stops and cannot be restarted until the fault is resolved. ISEE-CAM also operates on the fail-safe principle; when the camera connection is lost, the safety signal is lost and the machine stops.

A safety relay evaluates the signal from safety elements such as light curtains, two-hand controls, or emergency stops and ensures the machine stops safely. The connection should be designed according to the type of safety element used and the machine's stop chain. The fail-safe principle and relevant standards must be followed in circuit design.

In transfer press lines, since parts are automatically transferred, the safety system must work in coordination with the transfer mechanism. An independent safety zone is defined at each press and muting is applied during transfer. An AI camera system can reduce the need for muting since it can distinguish between the transfer mechanism and the human body.

EN 692 defines mechanical press safety requirements and EN 693 defines hydraulic press safety requirements. Both standards cover hazard analysis, protective device requirements, control system safety, and user manual topics. In Turkey, these standards have been transposed as TS EN 692 and TS EN 693.

According to the ISO 13855 formula: S = (K x T) + C. S: minimum mounting distance (mm), K: hand approach speed (2000 mm/s), T: total response time (sensor + control + press stopping time), C: additional distance based on detection capacity. For example, if T=0.2s and C=48mm, then S = (2000 x 0.2) + 48 = 448mm. Incorrect calculation creates a serious safety gap.

Light curtain bypass is a common and dangerous problem. Tamper-proof mounting, bypass detection sensors, and serious sanctions are required. The biggest advantage of AI camera systems like ISEE-CAM is that they cannot be bypassed through physical tampering; the system working through camera images continues without any physical connection being cut and bypass attempts are logged.

Safety device periodic inspections should be performed at least once a year; manufacturer recommendations may specify shorter intervals. Inspection covers sensor verification, response time measurement, and general operation tests. Daily operator checks should also be performed. ISEE-CAM's built-in diagnostics system instantly reports fault conditions, preventing safety gaps from forming.

To prevent forklift accidents, operator training, speed limitations, pedestrian-forklift separation, and technological solutions should be applied together. The most common accident types are pedestrian collisions, tip-overs, load falls, and ramp accidents. ISEE-CAM's forklift safety module detects personnel around forklifts with AI, provides collision warnings without requiring tags, and automatically stops the forklift when needed.

A forklift operator license is issued after theoretical and practical training at accredited training centers. Training duration is generally around 32 hours and it is recommended to be renewed at regular intervals. Operating without a licensed operator creates both an operational risk and increases the employer's administrative and legal liability.

As a general rule, the forklift speed limit should be 10 km/h indoors, 5 km/h in pedestrian-heavy areas, and lower on ramps and narrow corridors. AI camera systems like ISEE-CAM perform zone-based speed monitoring, detect forklifts exceeding the limit, and instantly warn the operator; speed violations are continuously and objectively kept under control.

A forklift pedestrian collision prevention system is technology that detects the distance between a forklift and a pedestrian and issues a warning or slows down the forklift when a collision risk arises. In UWB-based systems, each pedestrian must wear a tag; ISEE-CAM detects personnel from camera footage using AI and thus works tagless. This means visitors and personnel who forget to wear their tags are also protected.

For forklift-pedestrian safety in a warehouse, basic steps include marking pedestrian paths with yellow lines, placing mirrors and warning lights at intersection points, one-way traffic layouts, and speed limitations. However, physical measures alone are not sufficient. AI camera-based collision prevention systems continuously analyze pedestrian and forklift movements and provide real-time warnings when a collision risk arises.

UWB is a technology that measures precise location using radio signals. Distance is calculated using tags attached to forklifts and personnel; it is preferred particularly for proximity warnings in enclosed areas. However, it requires tag usage, battery maintenance, and anchor infrastructure. In sites with personnel who do not want to carry tags or with variable personnel flow, image-based AI systems can offer a more flexible alternative.

Forklift blind spots create serious risks, especially during reverse maneuvering and at intersections. Traditional solutions include convex mirrors, blue spot lights, and audible warnings. However, these measures are attention-dependent. ISEE-CAM's blind spot warning mode continuously scans the area around the forklift with AI and instantly warns the operator when it detects personnel in blind spots.

Working with personnel without forklift operator licenses is a serious OHS non-compliance. This situation can result in administrative sanctions during inspections; in case of an accident, the employer's legal and criminal liability increases. Licensed training, site rules, and technological safety measures should be applied together.

Electric forklift use is mandatory indoors; diesel forklift exhaust gases create serious health risks. Electric forklifts operate more quietly, so their approach may go unnoticed and pedestrian collision risk may increase. AI-based collision prevention systems are recommended for both types; ISEE-CAM works regardless of forklift type.

Forklift tip-overs occur due to excessive speed, sharp turns, unbalanced loads, sloped surfaces, and sudden braking. The stability triangle is formed by the two front wheels and the center of the rear axle; when the load's center of gravity moves outside this triangle, the forklift tips over. Operator training and speed control are the most effective measures.

Daily forklift inspection covers checking components such as brakes, steering, tires, forks, hydraulic system, warning lights, horn, mirrors, and seat belt. It should be completed by the operator at the beginning of each shift. AI safety systems do not replace daily checks but provide 24/7 assurance beyond the checklist by continuously monitoring operational safety.

A forklift fleet management system is a digital platform that centrally monitors forklift fleet usage data (operation time, speed, location, impact events, maintenance needs). ISEE-CAM works integrated with fleet management systems by adding safety events around the forklift to fleet data and provides a more comprehensive safety view.

Blue light projects a blue dot on the ground to notify pedestrians that a forklift is approaching; red zone light marks the hazardous area. These visual warnings aim to attract pedestrian attention but may be insufficient in noisy environments. They are not legally mandatory but are recommended as good practice. ISEE-CAM's audible alarm and automatic deceleration features provide active intervention beyond passive warnings.

Dock safety requires vehicle wheel chocks, dock levelers, vehicle restraints, adequate lighting, and communication protocols. The vehicle sliding or departing early while a forklift is entering can cause serious accidents. AI camera systems can monitor vehicle-personnel interaction at the dock area and detect and warn of unsafe conditions.

Forklift periodic inspections must be performed at least once a year by an authorized person according to the Work Equipment Regulation. Inspection covers structural integrity, hydraulic system, brakes, steering, lifting mechanism, and safety devices. Results are documented with a periodic inspection report. Continuous monitoring with technological safety systems is also recommended.

When AGVs and human-operated forklifts work in the same area, separated traffic paths, intersection management protocols, and mutual detection systems are required. AGVs detect obstacles with laser scanners but cannot distinguish between forklifts and humans. ISEE-CAM distinguishes between AGVs, forklifts, and pedestrians with AI, ensuring safe coordination in mixed traffic environments.

Zone-based speed limiting can be implemented with RFID tags, magnetic strips, or GPS-based systems. AI camera solutions like ISEE-CAM detect forklift speed through image analysis and log speed violations; speed limiting signals can be sent directly through PLC/SCADA integration.

Yes, a forklift accident is classified as a workplace accident under the relevant social security legislation and must be reported within 3 business days. If it is determined that necessary safety measures were not taken, administrative and criminal sanctions are applied to the employer. AI safety camera records play an important role in both documenting the accident and proving the measures taken.

Forklift safe driving training should consist of theoretical training, practical driving exercises, and field evaluation. Load handling rules, speed limits, pedestrian awareness, and blind spot management must be covered. Post-training monitoring of behavioral data from the field helps make operator awareness permanent.

Overhead crane safety fundamentals include not exceeding load capacity, regular periodic inspections, operator certification, sling selection, prohibition of working under the crane, and compliance with signaling rules. ISEE-CAM's crane safety mode continuously monitors personnel presence in the hazardous zone under the crane with AI and warns the operator in real time, enhancing the safety of lifting operations.

Crane periodic inspections should be performed every 3 months or once a year by an authorized engineer depending on their lifting capacity. TS EN 15011 and the Work Equipment Regulation are reference standards. Inspection covers structural integrity, ropes, hooks, brakes, and safety devices. Inspection documents must be kept for presentation during audits.

A tower crane operator license is issued after receiving theoretical and practical training from accredited training centers. Training covers crane mechanisms, load calculation, signaling, and safety rules. The license should be renewed every 5 years and the operator must undergo a health examination. AI camera-assisted personnel detection systems help the operator see risks in their blind spots.

A load moment limiter is a safety device that prevents the crane's lifting capacity from being exceeded. It continuously measures the load weight and crane arm angle, automatically stopping lifting in overload conditions. It plays a critical role in preventing tip-over accidents and is mandatory by regulation. AI camera systems work alongside moment limiters to also detect personnel presence under loads.

Crane accidents constitute a significant portion of fatal workplace accidents in Turkey. The most common causes are overloading, rope breakage, sling errors, wind effects, and operator errors. They occur most frequently in construction, port, and iron-steel sectors. The vast majority of these accidents are caused by personnel being present under loads and can be prevented with AI-based personnel detection systems.

An anti-collision system is technology that prevents multiple cranes working in the same area from colliding with each other. It monitors the position and speed data of cranes and performs automatic deceleration when a collision risk arises. Traditional systems are radar or laser-based. AI camera systems can detect both crane-crane and crane-personnel risks on a single platform.

A lift plan should include the load weight and dimensions, crane capacity chart, boom angle, ground bearing capacity, wind conditions, and sling angles. A hazardous area (exclusion zone) should be defined and personnel access should be controlled. AI camera systems support the safety elements of the lift plan by detecting personnel entering the defined zone.

Wind limits vary by crane type and capacity; generally, operations should be stopped above 60-72 km/h. Anemometers are mandatory on tower cranes and automatically alarm at preset values. Continuous monitoring is critical since wind conditions can change instantly.

In sling selection, the load weight, dimensions, center of gravity, sling angle, and SWL (Safe Working Load) value are the basic criteria. SWL must never be exceeded. Wire rope, chain, and polyester slings have different usage conditions. Damaged slings must be immediately removed from service and periodic inspections must be performed regularly.

In UWB-based systems, warnings are given by measuring the distance between tags worn by workers and the crane. However, UWB cannot protect personnel who do not wear tags. ISEE-CAM's crane safety mode detects all personnel under the crane taglessly from camera footage using AI; everyone is protected, including visitors and workers who forget to wear their tags.

Rope replacement criteria are defined in ISO 4309 standard. Criteria include the number of broken wires in a single lay (6) or total in 30 rope lays, more than 10% diameter loss, corrosion, and deformation. These criteria are inspected during periodic checks and hazardous ropes must be replaced immediately.

An exclusion zone is a prohibited area defined under and around the load during crane operations. It is marked with physical barriers, warning tape, and a signal person. However, physical measures may not always be sufficient. ISEE-CAM detects personnel entering the defined zone in real time, issues a warning, and sends a signal to the PLC to stop crane operations when needed.

OHS mode is a special operating mode added to crane control systems for safety purposes. In this mode, crane speed is reduced, movement is restricted, and safety functions are activated. It is used during maintenance and special operations. When integrated with AI camera systems, the crane can be automatically switched to OHS mode when a hazard is detected.

TS EN 15011 regulates the design and safety requirements for bridge and gantry cranes, while TS EN 14492-2 regulates safety standards for electric hoists and winches. These standards define requirements for structural strength, safety devices, control systems, and user manuals. Compliance with standards forms the foundation of crane safety.

Basic hand signals include: hoist (index finger rotating upward), lower (arm downward), travel right/left, stop (horizontal arm movement), and emergency stop (both arms crossed). Radio use is recommended in noisy environments. AI systems have the potential to automatically recognize hand signals in the future to provide crane control support.

In a crane accident, the employer bears criminal liability (negligent injury/death under criminal law) and civil liability (material-moral compensation) if it is determined that necessary safety measures were not taken. To avoid liability, periodic inspections, operator training, risk assessment, and technological safety measures must be fully implemented.

The overload protection system continuously measures weight with load cell sensors and automatically stops lifting when the nominal capacity is exceeded. Sensor calibration should generally be performed every 6 months or annually. AI camera systems work alongside overload protection to also monitor human presence under the crane, providing an additional safety layer.

Gantry cranes move on fixed rails; the safety risk is personnel in the rail area. Jib cranes rotate at a limited angle; load swing risk is high. Mobile cranes offer flexible placement but have the highest tip-over risk. In all crane types, an AI camera-based personnel detection system provides an effective common safety layer.

Modern cranes have IoT sensors that continuously monitor load, speed, wind, temperature, and operating hours and transmit data to a cloud platform. Maintenance tracking software enables pre-failure maintenance planning. AI systems like ISEE-CAM add visual safety data to IoT infrastructure, providing comprehensive monitoring from both operational and OHS perspectives.

Load drop root causes are grouped as sling errors, operator errors, mechanical failures, and environmental factors. The 5 Whys method should be used to uncover systematic issues underlying each factor. AI camera recordings document the moment of the accident and the preceding events with visual evidence, improving the accuracy of root cause analysis.

A collision prevention system is a safety technology that detects proximity between industrial vehicles and personnel and issues warnings or automatically stops the vehicle. Different technologies are used, including UWB (tag-based), LiDAR (laser distance), radar, and AI cameras (image analysis). ISEE-CAM provides tagless collision prevention by detecting personnel and vehicles from camera footage using AI.

UWB provides centimeter-level location accuracy; RFID has a shorter range and less accuracy; Bluetooth has low reliability in industrial environments. All three technologies require workers to wear tags. Their disadvantages include tag cost, battery maintenance, anchor infrastructure, and the inability to protect people who do not wear tags. AI camera systems perform visual detection without requiring any wearable device.

An AI camera system detects humans, forklifts, and obstacles in real time using deep learning and calculates collision risk. Its biggest difference from UWB is tagless operation. ISEE-CAM processes this technology with Edge AI directly on the camera, eliminating the need for a server. It also protects visitors and subcontractor workers who do not wear tags.

When making the technology choice, whether personnel will carry tags, infrastructure setup requirements, false alarm tolerance, integration capability, and the blind spot structure of the site should be evaluated together. While UWB-based systems provide high location accuracy, image-based systems can offer more flexible site management without requiring additional tags.

LiDAR measures the distance and shape of surrounding objects with high accuracy using laser beams. However, human-object discrimination is limited and performance may decrease in dusty/smoky environments. AI camera systems perform human-object discrimination with high accuracy through visual analysis, and ISEE-CAM's IP69K protection guarantees reliable operation in dusty environments.

Radar sensors are effective for speed and distance measurement and are minimally affected by weather conditions. However, reflections from metal surfaces, interference in confined spaces, and inadequacy in human-object discrimination can be issues. Alone, it is not sufficient for reliable collision prevention in a factory environment. AI camera systems eliminate these disadvantages by classifying objects.

Yes. ISEE-CAM connects directly to industrial automation systems via OPC-UA, Modbus TCP, and REST API protocols. When a hazard is detected, forklift deceleration/stopping, siren activation, and warning light activation are automatically triggered via PLC, and all safety events are reflected on SCADA screens.

If attaching tags to all workers is difficult (visitors, subcontractors, high turnover), a tagless AI camera solution is most suitable. On a single ISEE-CAM platform, you also get additional functions such as collision prevention, PPE detection, and zone monitoring. You can request a free discovery visit to get a solution proposal tailored to your facility.

Crane-crane collision prevention continuously monitors the positions of cranes operating in the same area and performs automatic deceleration or stopping when the safe distance is exceeded. Traditionally, distance sensors and PLC-based control are used. AI camera systems can manage both crane-crane and crane-personnel risks on a single system by monitoring crane positions through image analysis.

In UWB systems, anchor installation throughout the facility and tag distribution can take weeks. ISEE-CAM, being a standalone Edge AI device, can be quickly mounted without needing to stop production. It is commissioned by mounting the camera at a strategic point and connecting power; no server setup or complex infrastructure work is required.

In UWB systems, signal reflection from metal obstacles can cause incorrect position calculation. In AI camera systems, model quality and environmental conditions are determining factors. ISEE-CAM minimizes false alarms with models specifically trained for industrial environments and site-specific calibration; in case of false detection, an update is made within one business day.

In wearable device-based systems, workers carry UWB or RFID tags, vehicles have readers, and an alarm is given at a certain distance. The biggest problem is the inability to protect workers who forget to wear, refuse to wear, or whose tag battery dies. AI camera-based systems eliminate these issues by performing detection from images without requiring wearable devices.

ROI calculation considers prevented accident costs, insurance premium reductions, and productivity gains. The total cost of a single forklift-pedestrian collision accident can reach hundreds of thousands of dollars. Since ISEE-CAM offers multiple safety functions on a single platform, you do not need to purchase separate systems for each function and the ROI period is significantly shortened.

In UWB setup, anchors are generally installed every 15-30 meters throughout the facility, tags are distributed to each vehicle and worker, and the system is calibrated. In large facilities, hundreds of anchors and thousands of tags may be needed, creating high infrastructure costs and maintenance burden. In AI camera solutions, this infrastructure is completely eliminated.

Zone-based speed limiting defines different speed limits for different risk zones in the facility. ISEE-CAM can perform zone definition and speed monitoring with AI; through PLC integration, automatic speed limiting can be triggered when a forklift enters a specific zone.

RTLS is not yet legally mandatory in most jurisdictions. However, occupational safety laws obligate the employer to take all necessary measures to prevent occupational risks. At facilities where forklift-pedestrian collision risk is assessed as high, implementing technological collision prevention solutions can be considered within the employer's obligations.

Advanced systems operate in two stages: alarm in the warning zone, automatic deceleration or full stop in the danger zone. ISEE-CAM can perform automatic deceleration and stopping by sending signals directly to the forklift speed control system through PLC integration. This active intervention is much more effective compared to passive warning-only systems.

UWB systems are designed for indoor use. ISEE-CAM's IP69K protection rating and -20°C to +70°C operating range ensure reliable performance in the most challenging environments, including outdoors. It is unaffected by rain, dust, and extreme temperatures. Its detection range of up to 25 meters provides sufficient coverage for outdoor applications.

In UWB systems, anchor calibration, tag battery replacement, and firmware updates must be performed regularly. In AI camera systems, maintenance is much lower; lens cleaning and software updates are sufficient. ISEE-CAM's IP69K housing provides protection from dust and water, minimizing maintenance needs; software updates can be performed remotely.

A safety laser scanner is a device that detects surrounding objects with a rotating laser beam and produces a signal when entry into a defined safety zone is detected. It operates with a scanning angle of 180° or 270°. While reliable as a traditional sensor solution, it scans on a single plane. AI camera systems provide more comprehensive hazard detection through three-dimensional scene analysis.

When selecting a safety laser area scanner, scanning angle, protective field distance, resolution, response time, safety level, and integration interfaces should be evaluated together. Dust, vibration, and line-of-sight conditions of the application area are also determining factors in the selection.

When choosing a safety laser scanner, one should focus on application needs rather than brand. Protection area, warning area, PLC integration, muting requirements, ease of maintenance, and environmental conditions are the basic criteria for correct product selection. In areas requiring horizontal plane protection, laser scanners stand out, while camera-based solutions may excel in sites requiring more flexible visual analysis.

A laser scanner scans a wide area in the horizontal plane (180-270°), while a light curtain provides protection along a vertical line. Laser scanners are common in robot cells and AGVs, while light curtains are prevalent at press and machine entry points. AI camera solutions can cover both use cases on a single platform.

Safety laser scanners, light curtains, and safety mats are common options for robot cell safety. AI camera systems like ISEE-CAM detect personnel entering the robot work area from images, covering blind spots of traditional sensors and creating an additional safety layer.

Safety mats are used at press fronts, robot cell entrances, and hazardous area entries; they stop the machine by producing a signal when stepped on. Disadvantages include sensitivity to chip and liquid accumulation, risk of damage under heavy loads, and limited area. AI camera systems are not affected by these disadvantages since they do not require physical contact.

Muting is the temporary deactivation of the laser scanner's safety function under certain conditions (material passage, AGV movement). Incorrectly configured muting creates a serious safety gap. Since AI camera systems can distinguish between humans and materials through image analysis, they significantly reduce the need for muting.

Blanking is a function that allows the laser scanner or light curtain to ignore certain beam channels. Fixed obstacles or small objects are filtered. There are two types: fixed blanking and floating blanking. Incorrect configuration can create a safety gap.

The protection field is the zone that stops the machine when entry is detected, and the warning field is the zone that triggers an alarm when approach is detected. Zone dimensions are calculated based on machine hazard level, stopping time, and approach speed. Multiple field sets can be defined for different production scenarios using the scanner software.

On AGVs, a safety laser scanner is mounted facing the direction of travel to detect obstacles. Using dual scanners provides more comprehensive protection. ISEE-CAM distinguishes between AGVs, forklifts, and humans with AI through visual analysis and identifies the human-object difference that laser scanners cannot distinguish.

Safety laser scanners connect to safety PLCs via OSSD, PROFIsafe, or CIP Safety protocols. ISEE-CAM integrates directly with standard PLCs via OPC-UA and Modbus TCP protocols and with safety relays through dry contact outputs; no complex safety bus configuration is required.

SIL (IEC 62061) and PL (EN ISO 13849) define equivalent safety levels. Equivalences are SIL1≈PLb, SIL2≈PLd, SIL3≈PLe. The required level is determined based on risk assessment; SIL3/PLe is required for high-risk applications such as presses and robot cells.

IP65 is dust-tight and water jet protected, IP67 provides short-term immersion protection. IP65 is the minimum requirement for dusty environments. For heavy dust and washdown environments, IP69K is more suitable. ISEE-CAM's IP69K certification guarantees reliable operation even in the heaviest dusty environments; this is a protection level many laser scanners cannot achieve.

Outdoors, UV resistance, wide temperature range, IP protection class, and optical performance are critical factors. Many laser scanners are designed for indoor use and can be affected by sunlight. ISEE-CAM, with its -20°C to +70°C range and IP69K protection, is ideal for outdoor applications; its detection range of up to 25 meters provides wide coverage.

A safety edge sensor is a pressure-sensitive strip sensor mounted on the edge of automatic doors and moving machine parts that produces a signal upon contact. It prevents entrapment risks. It has a narrow application area and only responds at the moment of physical contact. AI camera systems provide proactive protection by detecting before contact occurs.

In press safety, a light curtain is the most common standard solution; it provides vertical protection. An area scanning sensor provides horizontal floor protection in front of the press. Using both together provides the most comprehensive protection. An AI camera press safety system monitors both vertical and horizontal zones with a single device and solves the false alarm issue with hand-material discrimination.

A 180° scan is sufficient for flat surface protection. A 270° scan can protect corners and L-shaped areas with a single device and provides wider coverage on AGVs. 270° models are generally higher priced. AI camera systems offer a wide field of view depending on lens and mounting angle.

Safety laser scanners typically have a detection range of 3-5.5 meters in the protection field and 20-40 meters in the warning field. SICK microScan3 reaches 5.5m in the protection field, and Leuze RSL 400 reaches 8.25m. ISEE-CAM's AI-based detection range of up to 25 meters provides advantages in wide area protection applications.

A light curtain can be selected for narrow entry points, and a laser scanner for wide floor areas. The most flexible solution is an AI camera safety system; with a single ISEE-CAM, you can monitor a wide area, distinguish between personnel, vehicles, and objects, add PPE control, and connect to your existing system with PLC/SCADA integration.

Yes, object detection models trained on fire and smoke datasets can provide fast detection and early warning.

Object detection, combined with pose estimation or activity recognition, can identify falls, slips, or abnormal events.

Yes, with high-resolution cameras and specialized models, hand positions can be tracked to prevent injuries.

Models detect and count people entering or present in hazardous areas, supporting evacuation and access control.

It provides real-time situational awareness, detects human and hazard presence, and supports automated incident reporting.

Yes, it is possible. Thanks to its flexible AI architecture, ISEE-CAM can be retrained to recognize objects or equipment specific to your business beyond standard objects (personnel, forklifts, etc.). Through dataset work conducted by our expert team, custom object definitions are made to the system, and tracking and inspection processes specific to your operations are quickly deployed.

Yes, secure integration with PLC systems can be achieved via OPC-UA, Modbus TCP, or REST API.

Detected events can be transmitted to SCADA screens as alarms, logs, and statistics.

In many facilities, safety measures rely on fixed sensors. In dynamic production sites, these systems create blind spots and risks are detected late.

Sensors are tied to fixed points and cannot cover all risk scenarios in variable site conditions.

Yes, but performance depends on camera quality and dataset diversity. Infrared or thermal cameras improve success in low-light conditions.

The most common challenges are occlusion, variable lighting, camera placement, and the need for models robust to real-world conditions.

Yes, ISEE-CAM is fully compatible with heavy industry conditions. According to its technical specifications, it provides uninterrupted performance at extreme operating temperatures between -20°C and +70°C.

Yes, ISEE-CAM is specifically manufactured for these challenging conditions. With its IP69K certification, it operates at full performance in heavily dusty areas and outdoors.

GDPR compliance in image processing systems is directly related to where and how data is processed. ISEE-CAM uses Edge AI technology that processes images directly on the camera without transferring them to a central server or cloud. This means personal data (facial features, identity-identifying details) never leaves the device and is used only for instant analysis (such as personnel/forklift detection) without being recorded. Consequently, the risk of data leakage is eliminated and data protection requirements are met at the highest technical level.

Since data is processed locally on the device without being sent to the cloud, privacy risks are significantly reduced.

Periodic data collection and retraining should be performed depending on environmental changes.

Changes in environmental conditions and data distribution drift cause performance degradation.

The system's hardware and software warranty is 1 year. Throughout the warranty period, our authorized experts immediately contact the customer to provide the necessary support for any technical situation. Both for the physical durability of the device and the continuity of the software, a fast and solution-oriented technical support process is maintained to ensure your operations are not disrupted.

In case a false violation (false positive) is detected in the system, collecting the erroneous data, retraining the AI, and updating and reactivating the system is completed within one business day. Thanks to our agile software infrastructure and expert team, site-specific optimizations are quickly carried out to ensure the system operates continuously with the highest accuracy rate.

Yes, the system has real-time status monitoring and error notification mechanisms. In case the camera shuts down, its connection is lost, or a hardware failure occurs, you can receive instant warnings through the dry contact outputs on the system or signals sent over the network. This way, you have the opportunity to respond to the situation without any interruption in your security and quality control processes.

It will take on a more central role integrated with autonomous systems, digital twins, and smart factories.

The combined use of image, LiDAR, thermal, and audio data increases detection accuracy.

Object detection is a computer vision technique that identifies and locates objects in an image or video, drawing bounding boxes around them.

YOLOv8, YOLO-NAS, Faster R-CNN, and SSD are currently widely used architectures for object detection.

TensorFlow, PyTorch, and Ultralytics YOLO are popular choices that support state-of-the-art object detection models.

High-quality, accurately labeled datasets with realistic scenarios are required for robust object detection in occupational health and safety applications.

Transformer-based models (e.g., DETR), self-supervised learning, edge AI deployment, and multimodal systems combined with IoT data are among the latest trends.

These systems analyze human presence and hazardous areas in real time and automatically stop the machine when a risk arises.

Data augmentation enables the model to learn more robustly across different lighting, angle, and scale conditions and improves generalization performance.

Class imbalance leads to low accuracy in detecting rare objects and should be balanced with special sampling techniques.

The most uncertain predictions of the model are selected, relabeled, and training efficiency is improved.

Real-time object detection generally requires GPU acceleration, edge AI devices (e.g., NVIDIA Jetson, Intel Movidius), or high-performance CPUs.

Edge devices, local servers, and cloud-based platforms are the main deployment options depending on latency and privacy requirements.

No, it is not mandatory. Unlike traditional systems, ISEE-CAM uses Edge AI technology that performs analysis directly on the device without needing to send images to an external cloud server. This way, your data does not leave the factory, full data privacy is ensured, and real-time results are produced within milliseconds unaffected by internet outages.

No, ISEE-CAM does not only report; it produces real-time outputs the moment it detects a hazard. Thanks to the digital outputs on the device, when a violation is detected, it can activate sirens, turn on warning lights, or immediately stop machine/forklift movement without requiring external software. All these processes occur within milliseconds thanks to Edge AI.

No, ISEE-CAM is a standalone system with its own processor. Since it performs all AI analyses on the camera (Edge AI), it does not require an external server, high-performance computer, or licensed software. Simply mounting the camera and providing power is sufficient for the system to operate.

Yes, you can use our system with your existing camera infrastructure. If you already have an installed system, we can integrate all your cameras into a central AI server to perform both OHS and quality control analyses. This way, you can transform your facility into an AI-powered smart production hub within seconds while preserving your existing hardware investment.

No. AI-powered cameras do not work by detecting heat like thermal cameras. AI cameras use visual data (RGB images) to detect humans, machines, movements, and hazardous zones. Through image processing and deep learning algorithms, they analyze objects in the scene based on shape, posture, movement, and context. Thermal cameras only detect temperature differences and can produce misleading results depending on ambient temperature, reflections, or weather conditions. AI-powered camera systems can detect human presence and risky interactions much more precisely and contextually without relying on heat; therefore, they offer a broader and more reliable usage area in industrial workplace safety applications.

Accuracy depends on the model, dataset quality, and environmental conditions. Models trained with relevant data and quality labels perform better in specific environments.

Regular model retraining, working with up-to-date data, improving labeling quality, and using advanced models reduce false positives.

Detected events can trigger alarms, notifications, or automatic stopping through API integrations or IoT devices.

In quality control processes, precision can be adjusted directly according to the manufacturer's needs and tolerance values. Beyond standard solutions, with project-specific high-resolution cameras and lenses, it is possible to work with high precision even at micrometer (micron) level details. The operating principle of the system is engineered to meet the most demanding quality criteria by optimizing with device selection tailored to needs.

Yes, it is possible to keep historical recordings through the system. Depending on your needs, you can store data on daily, weekly, or monthly periods by integrating a dedicated recording device into the system. Additionally, if internet connectivity is provided, you can redirect the images from the analyzing cameras to your own existing recording devices to keep the entire process on record.

Yes, the image processing camera can be monitored in real time. You can follow the system in real time either through a system-attached monitor placed near the camera or through a web interface over the network if internet connectivity is provided. This way, you can observe field operations live while AI analyses continue.

Yes, multiple cameras can communicate with each other and work in a coordinated manner. No high-speed internet connection is needed; communication is provided over a simple local network (switch connection) linking the cameras. This way, cameras can make joint decisions based on data from each other and produce real-time outputs accordingly, managing the safety of the entire site in a synchronized manner.

mAP is the key evaluation metric that summarizes the model's accuracy and recall performance across different threshold values.

IoU measures the overlap between the predicted box and the ground truth box to determine detection accuracy.

FPS indicates the number of frames the system can process per second, and in latency-sensitive applications, it directly determines performance.

Object detection enhances workplace safety by enabling automatic monitoring of hazardous areas, personal protective equipment (PPE) detection, and identification of unsafe behaviors.

AI systems provide businesses with concrete field data by analyzing risky behaviors and facilitate the development of a preventive safety culture.

Yes, object detection can reliably detect and track in real time whether workers are wearing hard hats.

Cameras with object detection identify and track personnel or vehicles entering restricted areas and issue warnings in case of violations.

Yes, object detection can be used in collision prevention systems by detecting forklifts, workers, and obstacles in real time.

Models are trained to recognize PPE equipment such as hard hats, vests, masks, and gloves, and automatic compliance checks are performed.

Yes, object detection can track the presence and location of safety signage in the workplace.

Object detection can be trained to recognize specific tools and detect stray or misplaced equipment in sensitive areas.

The biggest cause of workplace accidents in Turkey is humans and hazardous industrial machines working in the same area and the inability to detect risks in these areas early enough.

According to social security data, the most common workplace accident types are machine entrapment/entanglement, falls from height, object impact, forklift accidents, and electrical shock. Particularly in the manufacturing industry, limb entrapment in machines like presses and conveyors occurs at a high rate. The vast majority of these accidents result from the inability to timely detect human-machine interaction; AI-powered image analysis systems can detect these risks in real time and enable preventive intervention.

To reduce the accident rate, risk assessment, regular OHS training, engineering controls, and technology investments should be addressed together. Continuous monitoring of hazardous zones, PPE compliance tracking, and controlling human-machine interactions are critically important. AI camera systems like ISEE-CAM provide proactive safety across the factory with 22+ detection modes on a single hardware.

The law imposes obligations on employers including conducting risk assessments, appointing OHS specialists and workplace physicians, providing OHS training to employees, preparing emergency action plans, and reporting workplace accidents. Administrative fines are imposed when these obligations, which vary by hazard class, are not fulfilled. The law also requires employers to continuously update safety measures in line with technological developments.

According to social security data, hundreds of thousands of workplace accidents are recorded in Turkey every year and many workers lose their lives. Sectors such as manufacturing, metal, construction, and mining carry high risks. These data show that not only traditional measures but also technological and proactive safety approaches are needed.

Risk assessment consists of five steps: (1) Identifying hazards, (2) Analyzing and grading risks, (3) Determining control measures, (4) Implementing measures, (5) Monitoring and review. Methods such as Fine-Kinney, L-Matrix, or 5x5 matrix can be used. AI camera systems enhance the effectiveness of the process by enabling 24/7 automatic monitoring of hazardous zones identified in risk assessment.

A workplace accident must be reported to social security within 3 business days from the date it occurred. The report is made electronically through the official reporting system. Late reporting results in administrative fines for the employer. AI safety systems accelerate the reporting process by performing automatic recording and reporting at the time of the incident and contribute to the documentation process.

For a near-miss reporting system, a notification mechanism accessible to all employees should be established. Digital forms, mobile applications, or physical boxes can be used, and reported incidents should be classified and root cause analyzed. AI camera systems like ISEE-CAM automatically detect and record near-miss events, reducing dependence on human reporting and creating a more comprehensive database.

The ROI of OHS investment is calculated by comparing prevented accident costs (medical expenses, compensation, production losses, social security premium increases) with the investment cost. The average cost of a workplace accident can reach hundreds of thousands of dollars including direct and indirect expenses. Since ISEE-CAM covers 22+ different risk scenarios on a single platform, you do not need to purchase separate systems for each scenario and the total cost of ownership decreases.

For a safety culture, top management support, regular training, open communication channels, reward systems, and employee participation should be ensured. AI camera systems reduce the need for personal judgment-based inspections by detecting unsafe behaviors objectively and continuously; this strengthens the perception of fairness among employees and helps make the safety culture permanent.

Direct costs include medical expenses, social security premiums, compensation, and fines. Indirect costs include production downtime, personnel turnover, reputation loss, insurance premium increases, and management time loss. According to Heinrich's iceberg model, indirect costs can be 4-10 times the direct costs. Therefore, preventive technology investments like AI safety cameras that perform multiple detections on a single platform provide significant long-term cost advantages.

In very hazardous class workplaces, at least 40 minutes per month of OHS specialist work is required for every 10 employees. Workplaces with 50 or more employees in the very hazardous class must have a full-time OHS specialist. To increase the effectiveness of the OHS specialist, AI safety systems provide significant contribution; even when the specialist is not on-site, the camera system performs 24/7 hazardous area monitoring to maintain continuous risk detection.

Yes, it is mandatory. According to the law, in the very hazardous class at least once a year (16 hours), in the hazardous class once every 2 years (12 hours), and in the less hazardous class once every 3 years (8 hours). Training must also be provided before starting work and when changing jobs. Real field data and violation records obtained from AI camera systems make training content more concrete and increase employee awareness.

In a reactive approach, measures are taken after accidents occur, while in a proactive approach, risks are detected in advance and prevented before accidents happen. ISEE-CAM's AI-powered real-time detection feature is the technological application of truly proactive safety — it warns or stops the machine the moment a hazard occurs, without waiting for the accident to happen.

Root cause analysis uses methods such as 5 Whys, Fishbone (Ishikawa), or TRIPOD. In each method, the goal is to go beyond surface-level causes to uncover underlying systematic issues. AI camera recordings provide objective visual evidence of the accident moment and preceding events, making them an extremely valuable tool in root cause analysis.

To increase PPE usage, ergonomic and quality PPE procurement, regular training, and consistent inspection are necessary. However, continuous inspection by human eyes is not practical. ISEE-CAM's PPE detection mode automatically checks 24/7 whether hard hats, vests, gloves, and goggles are worn; it measurably increases PPE compliance rates by providing audible/visual warnings at the moment of detection.

According to social security statistics, the sectors with the highest accident rates are manufacturing (metal, automotive, food), construction, mining, logistics/warehousing, and energy. Press and crane accidents in the metalworking and iron-steel sectors, and forklift accidents in logistics stand out prominently. ISEE-CAM's multiple detection modes offer directly applicable solutions across all these sectors.

AI cameras, IoT sensors, and wearable devices are becoming widespread in industrial safety. Wearable sensors (UWB tags) provide location tracking but require a tag to be attached to each worker. AI camera systems work taglessly, detecting personnel and risks from images; they eliminate tag costs and maintenance burden while covering a much wider range of risks with a single device.

Before an audit, risk assessment reports, OHS training records, periodic inspection documents, emergency action plans, OHS committee minutes, and accident reports should be up to date. Machine guards, safety signs, and PPE usage should be checked. AI safety camera system records serve as strong evidence that concretely documents the business's proactive safety approach with data.

The OHS committee is mandatory at workplaces with 50 or more employees and meets 1-3 times per month depending on the hazard class. Agenda items include accident and near-miss analyses, risk assessment updates, training planning, employee complaints, and improvement suggestions. Violation reports and trend analyses obtained from AI camera systems provide a concrete data foundation for committee meetings, strengthening the decision-making process.

The most effective methods are hands-on training, real accident video analysis, simulation training, regular toolbox talks, and behavior-based safety programs. When unsafe behavior recordings detected by AI systems like ISEE-CAM are used as training material, they enable employees to see risks concretely and make awareness permanent.

According to social security statistical yearbooks, hundreds of thousands of workplace accidents are recorded in Turkey every year and thousands of workers lose their lives. It is also assessed that actual figures are higher due to unregistered employment. This picture clearly demonstrates the importance of investment in proactive safety technologies.

According to official social security data, approximately 1,500-2,000 workers lose their lives due to workplace accidents and occupational diseases in Turkey each year. Independent worker safety council data show higher figures. Turkey is among the worst-performing countries in Europe in this regard. AI safety systems have the potential to significantly reduce these losses.

According to social security data, the sectors with the highest workplace accident rates are manufacturing (metal, automotive, food), construction, mining, logistics/warehousing, energy, and agriculture. Machine entrapment in manufacturing, falls from height in construction, collapses in mining, and forklift accidents in logistics are prominently featured.

According to Eurostat and ILO data, Turkey is among the countries with the highest workplace accident mortality rates in Europe. These rates, which are well above the EU average, point to both deficiencies in regulation enforcement and inadequacies in technology use. AI-based proactive safety systems can play a critical role in closing this gap.

The most common accident types in the manufacturing industry are limb entrapment in machines (presses, lathes, conveyors), object impact, forklift accidents, falls, electrical shock, and chemical exposure. The vast majority of these accidents occur at human-machine interaction points and can be proactively prevented with ISEE-CAM's multiple detection modes.

Press-related workplace accidents are among the most significant causes of limb loss cases in Turkey. Hand-finger entrapment cases in eccentric and hydraulic presses are particularly high. According to social security data, thousands of press accidents are recorded every year. AI camera press safety systems have the capacity to prevent the vast majority of these accidents through hand detection.

Forklift accidents occur at high rates in warehouse, logistics, and manufacturing sectors in Turkey. The most common causes are pedestrian collisions, tip-overs, load drops, and ramp accidents. Collision accidents are generally caused by blind spots and inadequate pedestrian separation. ISEE-CAM's tagless forklift safety solution significantly reduces these accidents.

Crane accidents constitute a significant percentage of fatal workplace accidents in Turkey. They are concentrated in construction, shipyard, port, and iron-steel sectors. The most common causes of death are being crushed by falling loads, crane tip-overs, and falls from height. AI camera systems that detect personnel presence under cranes have the potential to prevent these deaths.

The total cost of a fatal workplace accident to the employer can reach millions of dollars. Direct costs include treatment, compensation, social security premium increases, fines, and legal fees. Indirect costs include production downtime, personnel turnover, training, reputation loss, and insurance increases. According to the Heinrich model, indirect costs are 4-10 times the direct costs.

The metal sector (iron-steel, aluminum, copper) is one of the sectors with the highest workplace accident rates in Turkey. The riskiest processes are press operations, casting, rolling, welding, turning, and grinding. High temperatures, heavy loads, and close contact with machines are the main risk factors. ISEE-CAM's IP69K protection and -20°C/+70°C range make it ideal for these harsh environments.

According to social security data, the most frequently injured areas are hands and fingers (especially in press accidents), feet, head area, eyes, and back. Hand and finger injuries are the most common in the manufacturing sector and can be significantly reduced with press safety systems and PPE. ISEE-CAM's hand detection and PPE detection modes are aimed at preventing these injuries.

According to social security data, the number of recorded workplace accidents has increased over the last 10 years; however, part of this increase is due to improvements in the recording system. The accident frequency rate relative to the number of workers has shown a fluctuating trend. The widespread adoption of technological safety solutions will play a key role in permanently reducing accident rates.

Worker safety councils maintain independent records through media scanning and field data, while social security only records officially reported cases. Due to unregistered employment, underreporting, and insufficient recording of occupational diseases, official data may not fully reflect reality. Both data sources should be evaluated together.

According to ILO data, workplace accidents and occupational diseases cost approximately 3.94% of global GDP. This amounts to trillions of dollars annually. Approximately 2.78 million people worldwide lose their lives due to workplace accidents and occupational diseases each year. Losses of this magnitude demonstrate the global importance of safety technology investments.

According to social security data, the 25-34 age group has the highest workplace accident rate; followed by the 35-44 age group. In young workers, inexperience and low risk perception, and in the middle-age group, long working hours and fatigue increase accident risk. AI safety systems protect all age groups without discrimination.

Workplaces in the very hazardous class (mining, metal, chemicals, construction) have workplace accident rates 3-5 times higher than the less hazardous class. OHS investments at workplaces in this class are critically important both as a legal obligation and an operational necessity. ISEE-CAM's IP69K protection and wide detection mode range are specifically suitable for very hazardous class workplaces.

The Occupational Health and Safety Law is the fundamental legal framework regulating the provision of occupational health and safety in Turkey. It applies to all workplaces, both public and private sector, and to all employees (including apprentices and interns). It imposes comprehensive obligations on employers including risk assessment, training, health surveillance, and emergency planning.

The employer is obligated to have risk assessments conducted, appoint OHS specialists and workplace physicians, provide OHS training to employees, prepare emergency action plans, report workplace accidents to social security, have periodic inspections conducted, and take necessary safety measures. Implementing safety solutions in line with technological developments is also within these obligations.

Administrative fines are imposed on employers who do not conduct risk assessments under the occupational safety law. Fine amounts are updated annually and can reach tens of thousands of dollars depending on the number of employees and hazard class. If the deficiency is not corrected, the fine continues to increase progressively.

The law requires OHS specialist appointment at all workplaces regardless of the number of employees. Workplaces with fewer than 50 employees in the less hazardous and hazardous classes can obtain this service from external OHS service providers. An OHS specialist is mandatory at all workplaces in the very hazardous class.

Workplaces are classified into three classes: less hazardous, hazardous, and very hazardous according to the Hazard Classes Communique. Metal processing, mining, construction, and chemical sectors are generally in the very hazardous class. The hazard class directly affects OHS specialist working hours, training periods, risk assessment renewal periods, and fine amounts.

A workplace accident must be reported to social security within 3 business days from the date it occurred. The report is made through the electronic reporting system. Late reporting incurs administrative fines. AI safety camera recordings play a valuable role as evidence in the reporting process.

An external OHS service provider (OSGB) is an organization that provides OHS services externally. Workplaces with fewer than 50 employees can obtain OHS specialist, workplace physician, and other health personnel services from these providers. Workplaces with 50 or more employees can establish their own OHS unit or use an external provider.

The current highlights in this regulation include periodic inspection, work equipment compliance, maintenance records, and safety measures to be taken during use. The employer is primarily responsible for keeping equipment in safe condition, documenting inspections, and appropriately informing employees.

Yes, CE marking is mandatory under the 2006/42/EC Machinery Directive. Using a machine without CE marking is a regulatory violation and increases the employer's criminal liability in case of a workplace accident. Administrative fines may be imposed. Safety retrofit solutions can be applied to older machines to ensure regulatory compliance.

Turkey has transposed the 2006/42/EC Machinery Directive into national legislation as the Machinery Safety Regulation as part of the EU harmonization process. All new machines placed on the market must bear CE marking and an EC declaration of conformity must be issued. Notified bodies perform type examination when required.

Yes, it is a legal obligation. In the very hazardous class, once a year (16 hours); in the hazardous class, once every 2 years (12 hours); and in the less hazardous class, once every 3 years (8 hours). Training is also mandatory before starting work, when changing jobs, and after more than 3 business days of absence. Training records must be documented.

Establishing an OHS committee is mandatory at all workplaces with 50 or more employees. The committee consists of the employer or representative, OHS specialist, workplace physician, employee representative, and human resources/personnel manager. It meets 1-3 times per month depending on hazard class and decisions are recorded in minutes.

A labor inspector reviews risk assessment reports, OHS training records, periodic inspection documents, emergency action plans, OHS committee minutes, workplace accident records, health surveillance results, and PPE delivery forms. Machine guards, safety signs, and the working environment are also physically inspected.

In a workplace accident, the employer may bear criminal liability under criminal law for negligent homicide (2-6 years imprisonment) and negligent injury (3 months-1 year imprisonment). Penalties are increased in cases of conscious negligence. Additionally, material-moral compensation and social security recourse lawsuits come into play. Safety technology investments serve as evidence that due diligence was exercised.

An OHS specialist is legally responsible for findings and recommendations within their scope of duty. Liability arises if they fail to report a detected risk or fail to inform the employer to take measures. However, when the employer does not follow recommendations, primary liability lies with the employer. The specialist can report this situation to the relevant authority.

Turkey continuously updates its machinery safety legislation within the scope of the Customs Union with the EU. The 2006/42/EC Machinery Directive has been transposed and new regulations in the EU are being followed. The harmonization process includes CE marking requirements, market surveillance, and adoption of technical standards.

Yes, periodic inspection of certain equipment is mandatory under the Work Equipment Regulation. Lifting and conveying equipment (cranes, forklifts), pressure vessels, electrical installations, elevators, and safety equipment are subject to inspection. Inspection intervals vary from 3 months to 1 year depending on equipment type.

Yes, preparing an emergency action plan is mandatory at all workplaces under the occupational safety law. The plan should include probable emergencies (fire, earthquake, explosion, chemical spill), evacuation routes, assembly points, assigned personnel, and contact information. Drills must be conducted at least once a year.

Updates to the law over time have focused on digital recording processes, remote working conditions, training, and notification applications. Current changes should be followed from official legislative sources; specifically, the Official Gazette and relevant government authority announcements should be referenced.