Just a few decades ago, workplace safety in manufacturing used to be an afterthought. Workers in the manufacturing industry may recall many near-miss accidents that have occurred, especially related to heavy equipment and lack of safety controls.
Video surveillance used to be the main source for controlling workplace accidents. Surveillance cameras were used to monitor the work environment and ensure safety. However, they were not ideal to help control the frequency of accidents. But, thanks to the advancement of Artificial intelligence and computer vision, we now have the power to create automated safety routines better than their human counterparts.
Now improved sensors can feed the algorithm with real-time data. They can collect data points, identify signals, and take corrective actions before an accident occurs.
For example, if a gas leakage is detected, increased temperatures will cause an alarm to inform the floor manager so that work can stop immediately, and the area is cleared.
The manufacturing work environment has always presented challenges, but AI has brought a major change to the safety scenario. By sensing the difference between what is safe and what is not, it can help create a more secure workplace.
Role of Computer Vision in the manufacturing work floor
Computer vision (CV) is an automated means to ‘watch while wandering’. Unlike a human supervisor who keeps a check on things while wandering around, computer vision can supervise not just one, but all areas on the work floor 24x7.
CV has the unique ability to keep a watch on the floor, see how workers operate heavy equipment and check the materials they are working with. It ensures that safety regulations are being followed by sending an alarm in the following situations:
- If a worker is not wearing the right safety gear – safety glasses, hard hat, etc
- If a machine is being operated without proper safety precautions
- If a worker is present in an area that is not considered safe
- If a person is standing behind a moving vehicle
- If a driver is going the wrong way on the work floor
- If a vehicle is parked in an area not designated for it
Computer vision identifies policy violation - spill detection in work floor
(Image source: Thingtrax)
Workplace safety solutions offered by Computer vision
Environmental safety – Since many facilities have CCTV cameras installed, computer vision can use the live feed to detect irregularities. If a dangerous situation is detected through live footage, the process can be stopped to avoid an accident. Multiple devices can be controlled, and the workplace can be made safer as it removes the need for workers to be present in dangerous areas.
Safety of materials – Scanners on production lines can be used to identify flaws in raw materials. They can identify the small defects that may escape the naked eye. Also, they could do this at a speed that is impossible for manual workers. CV can be used for multi-object detection that makes the production process more accurate.
PPE Enforcement – Personal Protective Equipment (PPE) protects workers against safety risks. PPE includes gloves, hairnets, hard hats, shoe covers, etc. They reduce the chances of injuries to workers and financial losses for the business. Enforcing PPE policy compliance has traditionally fallen upon managers and supervisors. But, with real-time object detection of images and videos, employees on the manufacturing floor can be checked for non-usage or incorrect usage of PPE without human help.
Computer vision detects the correct usage of PPE
Skyl.ai’s Computer vision implementation
Skyl.ai computer vision techniques help automate the safety monitoring processes in manufacturing plants. It can assist floor supervisors to ensure that all safety guidelines are being followed, boosting the production process and product quality. Our machine learning solutions help manufacturers prioritize other activities, reduce resource needs, and save costs.