COVID-19 has been the hot topic for us over the last months, but now it looks like we are returning to something like a new reality. The whole world is searching for ways to prepare for this. And, of course, employers need help to enable their employees to a safe work environment as possible.
​
Nowadays, healthcare workers rely on personal protective equipment to save themselves and their patients from being infected and infecting others. People have to use PPE in such places and conditions as they could not even imagine before. Public areas and transport, big corporations and banks, shopping malls and small stores, restaurants, gyms, barbershops, the police, and military facilities – there is a shortlist of areas to use PPE.

Special rules? Not really. Wearing more PPE every day? Makes sense!

Industry and governments have agreed to increase PPE manufacturing by 40% to meet rising global demand.

Safety is essential to any business. Different types of PPE are provided by employers across the world in many industries to protect employees from occupational health hazards.

In the last US Bureau of Labor Statistics report on Personal Protective Equipment, we can see that in most cases, when workers were injured, they were not using PPE. According to these statistics, 84% of workers with head injuries were not wearing hard hats, and 99% of workers who suffered facial injuries were not using face protection. ​

Many industries like manufacturing, construction, energy, etc., use machine learning models to engage their workforce and keep them safe. Today, organizations are doing a lot to improve the safe working environment in the Covid-19 pandemic.

Personal protective equipment: let’s explore the theoretical base

People who are working in different industries wear PPE equipment to minimize risks. The most common examples of PPE include respirators, gloves, aprons, fall protection, head, eye, foot protection, and full-body suits. A complete hazard control program that uses a variety of strategies to maintain a safe and healthy environment, not just using PPE. PPE does not reduce the hazard and does not guarantee total protection.
​
Hazards exist in every workplace, so technologies to protect workers are essential. The priority is to follow the hierarchy of hazards at their source or along the path between the source and the worker. It can include elimination, substitution, or engineering control.

Related: Human factors in safety control

The method to control a hazard at its source eliminates it from the workplace or isolates the worker from it. This approach may require the replacement of material with safety ones, isolation of hazards, and ventilation. Some safety features can be added to existing equipment, including redesigning the work processes or purchasing new equipment. Administrative controls such as work practices and education/training are also ways to control hazards.

Employees from all industries should be protected

PPE helps minimize or reduce the exposure or contact to harmful chemical, ergonomic, biological, or physical agents. Remember, the risk of injury will be reduced when PPE is used, but the hazard is not gone. For example, earplugs or headphones are appropriate for noise exposure. When the PPE is appropriately used, wearing hearing protection reduces the likelihood of hearing damage. Of course, using hearing protection does not eliminate the noise.
​
PPE should only be used:

• before controls are implemented as a temporary measure;
• in case of absence or unnecessary of other rules;
• during activities such as clean up, maintenance, and repair where pre-contact controls are not feasible or effective;
• during emergencies.

The main elements are:

• protection of workers
• compliance with applicable laws / regulations / standards / guidelines
• compliance with internal company requirements
• technical feasibility

An excellent comprehensive strategy considers the hazards, conducts a risk assessment, evaluates all possible control methods, integrates various approaches, and reexamines the controls frequently to ensure that the danger continues to be controlled.

How does it concern machine learning?

Now organizations and governments pay close attention to Environment, Health, and Safety (EHS) issues, putting a lot of effort into improving the sustainability of health and safety performances in the workplace.
EHS is a set of rules, processes, standards, and regulations developed to protect the environment, company employees, and society from any damage.

ML models are used in EHS for:

• Better data monitoring and efforts detection – such systems can alert if thresholds are exceeded, or some anomaly is detected. This allows compliance managers, engineers, or chemists to focus on problems rather than waste time checking areas where the ML-based system indicates as non-problematic.
• More effective predictive protection – ML-based systems can forecast failures or accidents and provide recommendations. It’s a significant opportunity in a risky environment. AI-based technology can improve water quality management, waste management, emissions monitoring, efficiency-enhancing, incident management, cost-reducing, and time and resources more efficiently.

Using ML to protect people in such industries as healthcare, sales, marketing, agriculture, manufacturing, logistics, automotive, etc.

Technological approaches to worker safety

AI and ML Computer Vision for workplace safety solutions – how does it work?

• Process and production safety – if the workplace has a CCTV system installed, a computer vision algorithm will detect abnormalities using CCTV’s live feed TV. Machine vision can automate production processes by controlling different or multiple devices. This removes the need for people to be in dangerous areas, increases efficiency, and makes the safe work environment.
• Vision-based detection and tracking – the deep learning-based object detection models use vision as the sensor for the machine to work in real-time. The search uses vision-based object detection and features like shape, color, motion, etc., of the object for the detection purpose.

For example, by using real-time video analysis and machine learning, such solutions can monitor employee safety. The systems can assess the suitability of workers’ protective equipment (PPE) by tracking entry and exit to the working environment. If an employee doesn’t comply with the requirements (forgets to put on headwear, glassware, and footwear), the system will signal, and access to the workplace will be denied till the person is properly equipped. So, it allows to avoid human errors and enhances higher employee safety in comparison with manual PPE checks.

The automation of PPE detection helps companies reduce the risk of accidents, fewer employer liability, provide analytics of workplace safety and improve operational efficiency.

Machine learning automation can collect and analyze data for the safety of the company employees or hospital staff. This allows them to focus on new, innovative and strategic ways to improve workplace safety instead of wasting time monitoring work sites manually. AI and Machine Learning change our lives in many good ways. Using cloud software, companies and healthcare organizations will be able to reduce injury and incident rates by over 50%.

​​It’s not just theory

Now responsible employers understand the importance of safety and have a vision about what it has to look like. It means that employees have to be on high alert to their surroundings, evaluate risks for the task, and take precautions appropriately. Algorithms can identify objects, edges, and velocity because of the advancements in machine vision. And it can do it faster and much more accurately than humans.

​The AI-based applications analyze the video stream from surveillance cameras, verify that employees are wearing any required protective equipment like masks/respirators, gloves, glasses, and medical gowns with 95% accuracy, recognize them, and report to the controller. The system can also control the body temperature of a person and track, keeping a safe distance in the company’s or healthcare organization’s building and territory.
​Those applications control the ones not ready to keep the rules. It can also track all contacts and persons in the risk group, which significantly decreases cases of infected staff of a clinic or a company.
The apps can be used for controlling healthcare or any company or institution (including schools and universities, industry, government, military facilities, and laboratories) to improve safety in the workplace.