Atlas Energy Solutions

John (JP) Srock

Connected Operations Steward

Safety cultures grow stronger when you recognize that incidents are signals of underlying risks. Our shift to integrating AI and operational technology into a unified safety ecosystem aims to empower everyone, fostering a sense of shared responsibility and confidence in safety improvements. We do not rely on a single dashboard or device, but on a connected platform that supports collective safety efforts.

Each system solves a different problem, but the real value comes from the flow of data between them. Vehicle telemetry, active driver coaching and video help us understand what happened on the road and in the moments leading up to an event and, in many cases, provide proactive coaching to change behavior, preventing the event in the first place.  

Proximity alerts show where people and equipment entered unsafe areas, actively stop the unit before an event occurs, and provide us with an opportunity to learn from these events through toolbox talks using real data. Site-monitoring tools add location, environmental, communication and alarm data. Frontline reporting and training tools capture what the crew saw, what conditions existed, what actions were taken, and whether followup learning occurred. When those inputs stay disconnected, safety is slow and reactive. When they are connected, safety becomes faster, more precise and more preventive.

That matters most after a near miss. Instead of waiting days for handwritten accounts, separate camera pulls and disconnected supervisor notes, we can reconstruct a more reliable timeline within hours. 

 Our shift to integrating AI and operational technology into a unified safety ecosystem aims to empower everyone, fostering a sense of shared responsibility and confidence in safety improvements. 

In one example, bringing these data sources together helped us determine that a single operator error did not cause repeated exposure, but rather a mix of congestion, visibility limitations and inconsistent pre-task communication. The response was not to blame. The process was redesigned: route changes, refreshed work rules, updated briefings, targeted coaching and clearer ownership of corrective actions. The result was a measurable reduction in repeat near misses of this type, faster incident review by investigation teams and a 62 percent reduction in closure time.

The AI component matters because it enhances our judgment through pattern recognition and prioritization. AI can identify risky behaviors and trends that might be missed manually, giving leadership confidence in early warning signals and proactive safety management.

We have also learned that safety technology only scales when workers trust it. If people experience these tools as surveillance, adoption stalls and the data loses its meaning. To address this, we will clearly explain what data is collected, why it is collected, who can access it and how it will be used; limit access; retain data for defined purposes; and keep all impacted employees in the loop for coaching and consequence decisions. Building this transparency ensures stakeholders understand and support the system, fostering trust and effective adoption.

The final lesson is that scaling is less about buying more tools and more about building operating discipline. Start with one or two high-risk workflows, establish a baseline and define a small set of shared metrics. Connect safety data directly into daily operational routines, not just monthly reports, to improve compliance and efficiency. When AI and connected systems are implemented this way, they do more than document risk; they enable the organization to see sooner, respond faster, and learn at enterprise scale.