How AI is helping reopen factory floors safely in a pandemic

It is becoming increasingly clear that Artificial Intelligence and Big Data will drive not only employee productivity, but also the safety of the workforce.

A worker at a factory in Tiruppur | Photo: Manisha Mondal | ThePrint

One of the biggest challenges post coronavirus lockdown has been to balance lives and livelihoods. How do factories and workplaces re-open while ensuring the safety of their employees, remains the pertinent question. As employers around the globe grapple with this, it has become evidently clear that the solution cannot be ‘one size fits all’. The way out needs a technology that could be adapted and fine-tuned to every factory floor, airport lounge and classroom. At the same time, it needs to be broad-based to meet international health and safety parameters.

In other words, the answer lies in adapting Artificial Intelligence and Internet of Things (IoT) technologies.

AI at work

For my team at BLP Industry.AI, the first step was to understand the practical difficulties that floor managers and supervisors in factories were facing, such as the inability to monitor their employees and if they were wearing the required safety gear constantly or not. Another difficulty was in ensuring social distancing not just among employees but visitors as well. Going through the inquiries we received from about 40 companies, both domestic and multinational, we learnt that some of them wanted their employees to submit a self-declaration document every day, which included questions on their health and whether they had visited a containment zone recently. Monitoring these daily self-declarations was proving to be cumbersome.

To ensure the safety of employees, an early warning system is necessary, so that anyone running high fever can be taken off the factory or office floor immediately. But there is no way companies can regularly monitor the temperature of every employee. Also, to prevent the spread of Covid-19, contact tracing is necessary, which again is a difficult task for employers. In addition, the companies would want to protect their supply chain, in particular the micro, small and medium scale (MSME) suppliers. Now, the employers wanted to achieve all of these and in a cost-effective manner.

We focussed on developing AI and IoT-based technology solutions for industry, educational institutions, hospitals, hotels, airports, etc., and came up with three broad ones that could be adapted based on the specific needs of different industries.

AI using camera feed, visual analytics

The first product, ‘Trust AI’, is a cloud-based solution that uses a combination of visual analytics, mathematical, and neural network models to analyse video feed. Any existing camera is connected to the cloud or the company’s server, which scans the feed in real time and immediately sends out an alert when a breach occurs. No new investment in CCTV cameras is required.

Alerts are sent to the safety officer or supervisors if safety gear usage (masks, helmets, safety jackets, etc.) or social distancing guidelines are not followed. In addition, the tool monitors hotspots in the factory and frequency of breaches so that managers can change the workforce on the floor.

Besides Covid-19, the technology can also be used in detecting fires, increasing workforce productivity, and reducing manufacturing defects. Institutions are also reducing security costs by replacing guards with computer vision models.

Mobile phone technology

The second product, ‘Us Pro’, is a cellphone technology meant only for enterprises and industries. It provides real time alerts on the employee’s cell phone when social distancing is breached. The phone sends out an instant alert to its owner, thereby providing an active defence system. Once an alert is triggered, it is recorded on the back-end AI application. This application, with relevant reports and dashboard, is only accessible to the health or safety officer of that particular factory or office. Privacy was a major factor for all the companies – and therefore a number of steps were taken such as limiting the use of technology only while the employee is at the factory or office, and keeping all communications between the phones encrypted. As a result, everyone remains anonymous.

The technology also tracks the employee’s temperature every few hours, and alerts the safety officer if there are signs of high fever. In case the person tests Covid positive, the AI application, using contact tracing, determines who among others has a higher probability of falling ill or contracting the virus.

Wearable devices

The third solution, ‘Spot AI’, involves wearable devices such as a wrist band or an ID card, which vibrates when a worker breaches social distancing and geo-fencing norms. 

The technology platform is normally used to drive operational and workforce productivity by locating and coordinating human, machine, and material flow on a factory floor. Increasingly, a number of US universities and Indian schools are evaluating the use of this technology as a non-intrusive way to create social distancing awareness.

Other sectors

These AI-based tools have helped companies be in a position to proactively implement safety measures in the workplace. Now, the retail sector too can use these applications to ensure that customers are adhering to safety norms such as wearing masks when in the store, or to keep a check on travellers at airports.

In crowded spaces such as offices and commercial buildings, these tools will help in protecting a large number of people if someone shows the symptoms of Covid-19. In most schools, cellphones are not allowed, so students can use wrist bands. The large Indian hotel chains that have properties across the country are evaluating a combination of these technologies. Hospitals, too, are evaluating the camera and wearable devices to keep their medical staff and patients safe.

The challenges

There were several issues that the partner companies and Industry.AI grappled with while developing these solutions. But the three that stood out were technology, privacy, and implementational challenges.

Based on regular feedback from partner companies on how to adapt the application to real-world requirements, a number of technology challenges were overcome.

Privacy was a critical issue that was extensively debated, and mitigating steps were taken. One, the technologies are only being applied within the confines of a factory, university, hospital, or an office. All the data remains anonymous, and only the safety and health officer or the appointed administrators of that particular company have access to them. Two, the camera feed is not stored on any server. All alerts are deleted after a certain period of time as per the company’s privacy policy.

As for implementation, the human resource managers and the culture of the company play a critical role. The implementation requires a good understanding of the workforce’s concerns and perspectives in order to ensure that the usage, scope and benefits of the technology are communicated clearly. Companies must create an environment of trust and convince their employees that it is in everyone’s best interest to adopt these preventive and safety measures.

Partnerships and the future

It is heartening to see corporations develop partnerships and come together in a time of crisis. In this case, the partnership between a few large corporations, supported by an able technology company, and subsequent pilot programmes with other industries, resulted in scalable and frugal solutions to be tested and implemented in a short period of time by a number of factories — from auto component major Lucas TVS to one of India’s largest electrical equipment companies, Havells. Moreover, some firms are now evaluating how these technologies can be used in a post-Covid world as well.

There are fears that the coronavirus pandemic may re-occur in waves, and the vaccines may not be ready for all virus mutations. This is forcing industries not only to partner with each other, but also with governments, to accelerate the adoption of next-generation technology. Given that supply chains have been broken and disrupted, we are seeing corporations accelerate their digital transformation plans to improve organisational productivity and decision-making.

As we move towards getting back to what we once considered ‘normal’, we will see the traditional paradigm being re-evaluated. And AI and Big Data will drive not only asset and employee productivity, but also the safety of the workforce.

Tejpreet Singh Chopra is the Founder and CEO of BLP Group, and former CEO of GE in India, Sri Lanka and Bangladesh. He is on the board of SRF, IEX, Anand Group, and AP Moller Maersk’s Pipavav port. He is a Young Global Leader of the World Economic Forum, and an Aspen Institute Fellow. Views are personal.