Subash Sharma
Waste management and recycling occupations are among the most dangerous in the world. It has been ranked the 6th most dangerous in the U.S. until recently, when it dropped to 7th. Even in sliding down the scale one notch, industry workers and the public are vulnerable to severe injury and fatality (SIF) risks. The constant risk management battle presents substantial threats to health, safety, and the environment (HSE). The industry walks the tightrope of whether to work faster or be safe, the former of which comes at a significant price tag. On average, organizations spend $40,000 per safety incident, which can balloon up to $1.2 million in the event of a fatality. As a result, leading waste management companies are now leveraging artificial intelligence (AI) to transform workplace safety.
Health and safety staff traditionally rely on lagging indicators for workplace safety. Computer vision AI transforms standard CCTV cameras into intelligent systems that detect nearly all leading indicators of accidents, revolutionizing safety practices. Armed with this data, teams can focus on proactively understanding and addressing root causes, mitigating risks, and preventing workplace accidents and injuries.
Case Example
Cory Group is one of the United Kingdom’s leading recycling and waste management companies. It ensures London and the South East have a safe, clean, and sustainable way of managing recyclable and non-recyclable wastes using its unique river-based infrastructure. Cory Group utilizes computer vision AI technology to manage nearly everything: traffic, monitoring and restricting hazardous areas, detecting unusual or risky behavior, detecting smoke and fire, and maintaining operational efficiencies.
Founded in 1896, Cory is a London-based recycling and waste management company that takes the city’s waste and turns it into low-carbon energy and construction materials. Cory Group receives around 900,000 tons of non-recyclable black bag waste a year, 89 % of which goes to Cory’s waste transfer stations via the River Thames.
The dry mixed recyclables (DMR) it receives, sorts, and segregates are for Western Riverside Waste Authority (WRWA) and its four boroughs: Hammersmith and Fulham, Kensington and Chelsea, Wandsworth, and Lambeth. In 2022, the Western Riverside, Wandsworth materials recycling facility separated 77,000 tons of DMR into 15 categories, including five different types of plastics, for onward recycling. They also manage two Household Waste and Recycling Centers for WRWA and the London Borough of Tower Hamlets.
The River Thames is Cory’s “green highway,” where non-recyclable waste is transported on non-self-propelled barges that use tugs for propulsion tugs via riverside transfer stations in Wandsworth, Battersea, the City of London, Tower Hamlets, and Barking and Dagenham. Using the Thames, Cory can save approximately 100,000 truck movements annually—a vital methodology for reducing road traffic for a safer, less polluted London.
Cory can process up to 850,000 tons of non-recyclable or residual waste a year at its Riverside 1 energy from waste (EfW) facility in Belvedere. The residual waste comes from local authorities and commercial and industrial businesses and is sent to the EfW facility via five river-based waste transfer stations. The Riverside 1 facility is one of the most extensive operational EfW facilities in the UK and the only one with a river infrastructure for receiving waste.
Given their business’s size, complexity, and risk profile, company executives turned to AI to implement and operate policies, manage risk, and deliver on their ongoing commitment to positively contributing to society.
Computer Vision AI – The Assistive HSE Technology
In 2022, Cory Group engaged with U.S-based SparkCognition to deploy and implement Visual AI Advisor, a computer vision solution designed to transform existing CCTV camera infrastructure into an intelligent visual analytics and alerts platform. The computer vision AI technology analyzes video feeds in real-time, identifying everything—critical safety measures, security, visual inspections, productivity, and situational awareness problems—and then reports on what it “sees.”
The program was fully implemented at Cory Group and operational by the end of 2023 and included 13 sites and tugs on 145 cameras. The solution was customized for Cory to categorize events and scenarios and then log them – and included safe events as well as those that are risky. For potential high-severity events, the solution was designed to immediately notify company authorities at each location via email, SMS, and WhatsApp. It was programmed for specific and exhaustive scenarios and use cases, including:
- Traffic management
- Vehicle/forklift speeding
- Man/vehicle/forklift interface
- Traffic rules compliance
2. Restricted or hazardous areas
- A person under suspended load/lifting equipment
- Machine guarding
- Monitoring safety barriers
- A person in a restricted area
3. Hazardous situations
- Dropped tools/objects
- Monitoring prohibited/hazardous materials
- Monitoring spills
4. Behavioral situations
- Man down
- Slips/falls
- Unusual behaviors such as climbing and running
- Human ergonomics
- Unauthorized use of mobile phones
5. PPE
- Monitoring the use of all PPE, such as safety helmets, vests,
- Safety harness when working at a height
6. Fire and Smoke
- Visible flame/smoke
- Person smoking
- Blocking of fire equipment/fire exits
7. Operational Efficiency
- Que management
- Dwell time
- Monitoring of debris on the floor for cleaning
8. Security
- Intrusion
In practice, computer vision AI technology detects anything unusual or unsafe, such as a worker in the path of a moving vehicle, someone working under a suspended load or within proximity to a hazardous situation, an unconscious or prone individual, a fall or stumble, the absence of a specified PPE, and even smoke, open flames, or someone smoking. The technology then acts upon the scenario by quickly alerting the appropriate stakeholders, helping to prevent a workplace accident.
It’s early in its deployment, but since December 2023, the AI technology has caught true positives. Cory Group anticipates this AI technology will be integral in shaping its HSE strategy.
Conclusion
Any organization with zero-incident goals for its workplace understands that reactionary workplace safety practices and manual or semi-automated safety incident management processes are grossly insufficient. Forward-thinking HSE managers are embracing proactive approaches designed to generate awareness of hazardous conditions and reduce the number of unsafe acts.
AI uniquely empowers HSE managers by harnessing advanced computer vision and other AI technologies to proactively identify and prevent unsafe acts, near misses, and hazardous conditions. Companies benefit by reduced overall costs due to fewer incidents and increased worker morale. It is genuinely a paradigm-shifting technology that HSE managers can use to create a system of improvements across their entire operating environment while increasing productivity.