Visual AI for HSE: Creating a System of Improvement

demo

While visual AI is effective for many different applications across many different types of organizations, leveraging visual AI for HSE (Health, Safety, and Environment) to identify activities that impact safer workplace environments is an especially critical use case.

The need for better HSE solutions today is tremendous: despite ongoing efforts to minimize accidents, injuries, and deaths in the workplace, there were over 4.2M work-related medically consulted injuries in 2021 and 4,472 preventable work deaths. What more can we do to proactively identify and address the root causes of these safety incidents to prevent them in the future?

SparkCognition Visual AI Advisor enables HSE managers to quickly and reliably address procedures and behaviors causing risky situations, create a system of improvement, ease the safety burden on their employees, and simplify their HSE program improvements.

Read our use case: Preventing Forklift Accidents With Visual AI

Proactive HSE Starts with Visual AI

SparkCognition Visual AI Advisor enables HSE managers to automate HSE monitoring, analytics, and real-time notifications. Visual AI Advisor can autonomously interpret the action wherever cameras can see, opening up powerful opportunities to optimize safety practices and increase the value of a factory or warehouse environment’s camera infrastructure.

That’s because any company, factory, warehouse, etc., with existing CCTV cameras can rapidly implement Visual AI Advisor to generate actionable analytics and real-time alerts—transforming their underutilized video feeds into a dynamic, proactive visual AI platform. Visual AI Advisor scales to thousands of cameras in days, utilizing widely available and inexpensive GPU nodes. It works with virtually any camera system installed on a site, including CCTV, PTZ, drones, and mobile devices, consolidating all video feeds into a single interface. With a low-code/no-code integration framework (no data scientist or programmer required), Visual AI Advisor deploys either on-premise or off and keeps data secure on-site or in the cloud. And its continuous-learning training module improves product performance over time to mitigate false positives and false negatives.

How Visual AI Advisor helps managers investigate root causes of near-misses and unsafe acts

A traditional safety incident management process today begins with a verbal or written report from a worker or manager on the factory floor—assuming it is a serious enough incident that they would actually stop work, gather all the details, and report up the chain of command using whatever tools/protocols they have in place. Unfortunately, this might be after a serious accident where someone got hurt, work is disrupted, and bottom-line costs are impacted.

Some parts of this process may be semi-automated, but typically, the process relies on manual human intervention to move forward. Yet how consistently would a worker with a long day of work ahead of them pause and make a note of a safety violation? Would it affect their motivation to do so if they were concerned that their boss would admonish them or not understand the circumstances? Or what if they don’t want to get a favorite co-worker written up?

Thus, a large and crucial part of the burden falls on employees to self-report and stay involved through a potentially lengthy, after-the-fact investigation process. That type of process is prone to human error and negligence.

Supervision, risk assessments, and process auditing are strenuous and time-consuming for HSE managers trying to overcome this flawed paradigm. 70% of HSE managers surveyed by SparkCognition said that training, controls, and procedure compliance were their biggest challenges.

Now contrast the scenario described above to our proactive solution leveraging computer vision—one that automates the recording of near misses and unsafe acts, sends real-time alerts to prevent accidents and reduce alert fatigue, and allows HSE managers to specify unsafe actions that should shut down a process.

Using the cameras already installed on-site, SparkCognition Visual AI Advisor automatically records unsafe incidents in real-time using CCTV cameras found in most workplaces. Data flows to robust dashboards where supervisors can easily track trends and discover root causes without depending on a human initiating or participating in its process.

Visual AI Advisor tracks incidents from multiple streams into a single dashboard, where all the information can be filtered by type, facility, and time of day with contextualizing snapshot images—reducing investigation time and aiding root-cause analysis. Report-building features quantify the incident reduction efficacy of new controls and procedures.

Visual AI Advisor’s capacity to identify, analyze, and alert to near-misses and unsafe acts in real time, 24 hours a day—leveraging accuracy and efficiency beyond human ability—enables HSE managers to:

  • Observe and address previously undetected warning trends, patterns, and isolated incidents.
  • Update training and controls to target patterns.
  • Measure and quantify the success rate of their changes.

 

Using Visual AI to optimize toward a zero-incident workplace

Keep in mind that while many, if not most, companies today already have CCTV cameras installed (which may or may not collect and store visual data in some capacity), the data produced by the cameras is less than actionable. Playback is retrospective, and it’s not unusual for data to be deleted within a short time frame. If the playback is monitored live at all, it’s unrealistic, if not impossible, for humans to watch all the action on every camera to identify safety incidents thoroughly around the clock. Studies have shown that humans performing the task of watching video feeds struggle to pay attention for more than 20 minutes and engagement plummets rapidly when more than one camera competes for attention.

But these same companies can use SparkCognition Visual AI Advisor to extend the value of their existing cameras to provide active monitoring of defined HSE violations or other relevant data points and send real-time alerts to designated supervisors or HSE managers, so they can take immediate corrective action, as needed.

By automatically tracking target KPIs in a single dashboard, HSE managers will now have comprehensive data with which they can create and measure a meaningful system of improvement to optimize toward a zero-incident workplace.

Organizations can automate visual analytics on multiple real-time video streams at once. HSE managers can target and collect essential data on performance, people, and assets from their configurable, drag-and-drop, web-based and mobile app dashboards. They’ll get role-based alert notifications immediately when safety non-compliance, security breaches, customer experience signals, etc., arise, delivered via their web dashboard, mobile app, email, SMS, WhatsApp, and/or on-site alarms.

In all of these ways, Visual AI Advisor reduces employees’ safety burden and simplifies organizations’ HSE improvement programs by leveraging computer vision-enabled CCTV cameras to automate information gathering and real-time notification.

Visual AI Advisor supports over 125 end-to-end computer vision use cases for safety, security, visual inspection, productivity, and situational awareness. An HSE manager may opt to use a subset of these use cases to target a workplace safety issue, like forklift accidents, using the following use cases:

  • Object Detection & Tracking (vehicle speed)
  • Scene Understanding (turnaround time)
  • Facial Recognition (authorized personnel)
  • Activity, Event & Threat Detection (collision risk).

 

Organizations can optimally configure Visual AI Advisor for their needs and environment, whether they are construction sites, retail stores, manufacturing floors, mining operations, or oil platforms. They can flag HSE compliance failures or insufficient personal protection equipment, identify persons in distress, spot potential damage to inventory, impending accidents, and more—resulting in cost savings, improved reputation, and increased overall safety in daily operations.

While the development of computer vision and visual AI began decades ago, in recent years, the technological maturity and adoption of such systems have matured rapidly. SparkCognition Visual AI Advisor minimizes time-to-value for operationalizing enterprise computer vision technology, delivering sustainable benefits for a wide range of industries across important business categories, including HSE.

To learn more, read our use case: Preventing Forklift Accidents With Visual AI

 

Latest blogs

Sunny side up NERC compliance hero image
Blog
Campbell LeFlore

Stay Sunny Side Up on NERC Compliance

Watch our webinar: NERC Compliance: Best Practices in Solar GADS Reporting In the good news department, renewable electricity capacity is continuing to grow by leaps

Read More

SparkCognition is committed to compliance with applicable privacy laws, including GDPR, and we provide related assurances in our contractual commitments. Click here to review our Cookie & Privacy Policy.