Despite endless efforts to minimize workplace accidents, injuries, and deaths, there were more than 2.8M work-related injury and illness cases in 2022. Workplace fatalities increased by 7.5 percent from 2021 to 5,486.
Farming, fishing, and forestry had the highest fatality rates, with construction and extraction jobs a close second place. What more can we do to mitigate this trend and proactively identify and address the root causes of safety incidents, injuries, and deaths in the workplace? Can artificial intelligence in the workplace be the solution?
Visual AI technology is effective for many different applications across a wide range of industries, but leveraging its capabilities for Health, Safety, and Environment (HSE) to create safer workplaces is an especially critical AI use case.
The best visual AI technology on the market enables HSE managers to quickly and accurately address processes, behaviors, and conditions that cause unacceptable risk. Companies then can improve these situations in a targeted manner, easing the safety burden on employees. This enables organizations to achieve their HSE program goals.
Manual HSE processes are reactive
Typical safety incident management processes begin with a verbal or written report from a factory worker or manager—assuming it’s a serious enough incident to have actually stopped work. Information is gathered about the details of the event and then reported up the chain of command using whatever tools/protocols are in place. This is a reactive approach to worker safety.
In some organizations, elements of this process are semi-automated, but the process usually relies on human intervention to work its way through the incident-management process. A large and crucial part of the safety awareness burden falls to employees to self-report and remain involved throughout a potentially time-consuming, after-the-fact investigation process. It’s a process almost guaranteed to suffer from human error.
Supervision, risk assessments, and process auditing are time-consuming for HSE managers. SparkCognition recently surveyed several such managers about this critical challenge and 70% said that training, controls, and process compliance were their biggest obstacles to success.
Contrast this scenario with a proactive solution that leverages computer vision enabled by an existing camera infrastructure—one that automatically records hazardous conditions, near misses, and unsafe acts, sends real-time alerts to workers that prevent accidents, and informs managers of unsafe conditions/actions that should shut down a process or cause it to be modified.
Using existing on-site cameras, visual AI tools automatically record unsafe conditions and incidents in real-time. This visual information then flows to dashboards where supervisors can easily track trends and identify root causes without relying on manual intervention. Multiple streams of activity can be integrated into a single dashboard, where all the information is then filtered by type, facility, and time of day, including contextualizing snapshot images. This reduces investigation time and aids root-cause analysis. Report-building features also quantify incident reduction efficacy for all updated controls and procedures.
The best visual AI products enable HSE managers to:
- Observe and address previously undetected trends, patterns, and isolated incidents.
- Update training and controls to target problematic patterns.
- Measure the success of process or environmental changes.
Visual AI optimizes safety, enabling zero-incident workplaces
Many companies today already have significant CCTV camera infrastructures (which may or may not collect and store visual data). But frequently, the data produced by these cameras is less than actionable. Playback is retrospective, and it’s common practice for data to be deleted soon after it’s been collected. Even in instances where video is monitored live, it’s unrealistic to expect humans to accurately view all the action from every camera and to identify safety incidents thoroughly and without error. Studies have shown that humans performing the task of watching video feeds struggle to pay attention for more than 20 minutes and engagement falls off rapidly with multiple cameras.
The promise of visual AI is active monitoring of defined HSE violations or other relevant data points and sending real-time alerts to managers, supervisors, or even workers themselves so that they can take immediate corrective action.
By automatically tracking target KPIs in a single integrated dashboard, HSE managers now have access to comprehensive safety data, empowering them to create a measurable improvement system that can lead to a zero-incident workplace. Users can automate visual analytics on multiple real-time video streams at once, targeting and collecting essential data on performance, people, and assets from configurable, drag-and-drop, web-based and mobile app dashboards. They’ll receive role-based alert notifications immediately whenever safety compliance processes are violated or security is breached, all delivered to the web dashboard, mobile app, email, SMS, WhatsApp, and/or via on-site alarms.
Visual AI Advisor supports over 125 end-to-end computer vision use cases for safety, security, visual inspection, productivity, and situational awareness. HSE managers can choose a subset of these use cases to target a specific workplace safety issue, e.g., forklift accidents, by using some or all of the following use cases:
- Object Detection & Tracking (Are vehicles moving too quickly or erratically?)
- Scene Context/Understanding (Which area of the facility are we viewing?)
- Facial Recognition (Are the individuals trained and authorized to be in this area? Are they wearing proper PPE?)
- Activity, Event & Threat Detection (Is there a collision risk from vehicles being too close to people or fixed objects?).
Organizations have the flexibility to configure Visual AI Advisor to best suit their processes, environment, and infrastructure whether construction sites, retail stores, manufacturing floors, mining operations, or oil platforms. They can generate alerts for HSE compliance failures that include insufficient personal protection equipment, persons in distress, potential damage to inventory, impending accidents, and more—resulting in cost savings, improved reputation, and increased operational safety.
Continuing year-on-year increases in workplace risk and worker accident rates do not have to be accepted as a normal state of affairs. The development of computer vision and visual AI, which began decades ago, has in recent years achieved a level of technological maturity and adoption that such systems can now deliver genuine value to enterprises. SparkCognition Visual AI Advisor minimizes time-to-value for operationalizing enterprise computer vision technology, delivering sustainable HSE benefits across a wide range of industries and operational processes.
SparkCognition Visual AI Advisor (VAIA) enables managers to automatically monitor HSE status, analyze risks as (or even before) they emerge, and deliver real-time notifications to team members that reduce or eliminate these risks. VAIA autonomously sees and interprets employee actions and environmental conditions wherever cameras can see, creating powerful opportunities to enhance safety practices and thus increase the value of a factory or warehouse camera infrastructure.
Learn how SparkCognition Visual AI Advisor reduces HSE risk and improves worker safety, then check out the case study Preventing Forklift Accidents With Visual AI.
