How to leverage AI computer vision to manage supply chain EHS risks

We have been interviewing our clients from a variety of industries ranging from Oil and Gas, automotive, FMCG to pharmaceuticals. When asked for the biggest EHS pains in their workplaces, the vast majority rank issues in packaging, warehousing and logistics in the top three. Here are a few reasons why Supply Chain EHS issues manifest themselves as one of the highest risks across industries:

Loss of control and reduced transparency due to outsourcing specialized activities to multiple vendors and subcontractors creating a complex chain of organizations across broad geographies.
Underreporting of incidents exacerbated by self-reporting biases result in poor decision-making and inaccurate evaluations of predictability of future incidents.
Temporary contingent workforce is employed with high turnover rates.
Inherently higher risks associated with the activities such as line of fire, man/machine interaction, road transport, traffic management, manual handling, working at height, etc.
The factor of uncertainty due to unknown contextual risks compounded with the dynamic nature of supply chain management such as supply disruptions, financial, legal and scope related schedule changes resulting in last minute changes at the frontline, causing workplace injuries.

When we dig deeper into the warehousing and storage industry, here is the trending with some data and facts¹:

In the US, the number of warehouse employees has doubled in the last ten years. The number of warehouses in the US only has increased by 20% between 2013 and 2018.
Warehouse workplace deaths are the highest they’ve been in 12 years with 213,100 warehouse-related injuries and 8,300 ware house-related illnesses reported in 2020 alone.
In 2019, the warehouse sector in the US saw the second-highest number of fatal work injuries out of any industry. In the UK, the transportation and storage industry has the fourth largest number of fatal injuries in 2022.

Here are the top-5 high risk hazards type in the warehousing and storage industry:

Trips, slips and falls: In a warehouse environment, trips, slips and falls can often occur due to the dynamic nature of this type of task. Hazards such as waste boxes, pallets, cables or other substances can all cause incidents. Failure to identify or remove a hazard is one of the key contributing factors. According to the U.S. Department of Labor, tripping, falling and slipping make up most of what is called “general industry accidents.” Slip and fall accidents make up 15 percent of all accidental deaths, 25 percent of all injury claims and 95 million lost work days each year.
Crush machinery injuries including Forklift injuries: In a warehouse, the most common crush injuries result from powered operational tools such as forklift, trucks and machinery. These types of incidents can be severe, especially if workers get caught in between or trapped under a tripping machinery. Forklifts are one of the biggest hazards in a warehouse. OSHA estimates that the number of forklift-related accidents is nearly 100,000 per year; around 20,000 workers in the US are injured in forklift accidents alone each year, and 25 percent of those accidents happen when a forklift overturns. 90 people are killed in those forklift accidents each year. Some other interesting facts to be considered are as follows:

90% of forklifts are involved in some type of accident over the course of their lifetime. (It is estimated that there are 856000 forklifts in the US alone.)
11% of forklifts in the U.S. are involved in an accident each year.
80% of forklift accidents involve a pedestrian.

Injuries caused by overexertion: Manual handling, lifting or carrying cases constitute 20% of non-fatal injuries in the workplace. In a warehouse, workers often have to lift heavy items.
Falling objects or dropped objects including collapsing racks: Being struck by a falling object is another common warehouse injury, which constitutes around 10% of non-fatal injuries in the workplace in 2018/19 that are the result of a moving object.
Falling from height: In a warehouse, falling from height cases usually occur because of using improper tools or equipment such as using forks to climb up, climbing up the racks above a certain height without a proper ladder or a Mobile Elevating Working Platform, etc.

Companies explore ways to adopt digital technology to manage these hazards more proactively. Future of the supply chain is shaped from now, and as intenseye, we are working with our customers to save lives at their workplaces and prevent life-altering injuries through AI computer vision.

As intenseye, we address 80% of these top hazards through our pre-trained, off-the-shelf, easy-to-deploy EHS use cases with high accuracy rates. Moreover, intenseye can help manage these risks in a proactive way, instead of relying on lagging indicators:

Intenseye will identify the unseen hazards, help to understand the severity and frequency of those and enhance quality of communication through real-time images, video alerts, analytics and task management tools.
Intenseye will improve visibility to the third party sites located in other countries or remote workplaces, which will close the gap between shop floor and corporate offices where the key decision makers are empowered with real-time data. The contractor management process will be underpinned by comprehensive analytics.
Intenseye will prevent the self-reporting biases and encourage the higher tier vendors to be more transparent with their concerns, which leads to higher quality data and better risk management. It will also demonstrate the effectiveness of the corrective actions in a more quantitative way while fostering cross-learning opportunities across various departments, facilities and vendors.
Intenseye enables organizations to design better interventions through tailoring to the local context, especially in Emerging Market (EM) suppliers, which will improve the working conditions of the workforce in line with UN’s Sustainable Development Goals (notably, SDG 8) while creating long-lasting social impact.

Here is just an example from one of our global clients. Using our dynamic delimitation AI model, they manage to reduce forklift-pedestrian interaction by over 75% in one month, significantly reducing the risk of an accident.