How Co-Innovation Helped Advance Serious Event Prevention with the PSI AI Advisor
Serious-event prevention depends on recognizing risk before it becomes an incident. For Sandra Neale of GE Vernova, that means learning from the safety concerns, stop-work reports, incident descriptions, and other information teams already collect across their operations.
Around early 2020, Sandra began working with the Benchmark Gensuite team on the PSI AI Advisor. The collaboration developed alongside GE’s work to strengthen its focus on potentially significant events and identify the conditions that could precede a fatality or other serious outcome.
In this Subscriber Spotlight, Sandra explains how the relationship moved beyond software implementation. GE contributed practical EHS knowledge and operational experience, while Benchmark Gensuite brought technical expertise, product-development support, and perspectives from its broader subscriber community. Together, the teams explored how AI could help professionals find meaningful signals in their safety data and act earlier.
Key Takeaways from Sandra Neale
- Collaborative innovation connects subscriber experience with product development, helping turn operational ideas into capabilities that support real EHS work.
- The quality of safety descriptions directly affects the quality of AI-supported analysis. Better inputs help teams identify potentially significant events more consistently.
- Connected safety data can help professionals narrow their focus, recognize repeated events, and identify precursors before a serious outcome occurs.
- Organizations can begin with a focused set of applications and expand in phases as the workflows demonstrate value.
How did your work to identify potentially significant events more proactively begin?
Sandra: I spent quite a bit of time around early 2020 working with Natasha and the team on what we call the PSI AI Advisor. It was a really fun project to work on with the digital technology team at Benchmark Gensuite and identify potentially significant events that can help us prevent fatalities.
We were on that journey at the same time, focusing our energy and actions on those weak signals and potentially severe events. It then moved into this advisory role on the AI side.
One of our other GE colleagues generated the idea for Describe-It AI. We started using it to improve the quality of the descriptions. Because we were one of the first companies working on the PSI AI Advisor, we found that the better the descriptions were, the better the advisor worked to identify potentially significant events. That gave us a deeper look at being more proactive in finding some of those potential events.
It is really about identifying or predicting where the next potentially significant event or fatality could be in your operation.
How can AI help teams find the safety signals that require attention?
Sandra: Think of all the data we have in the system. We can look at the precursors, stop-work reports, and concerns that are to the left of boom. Nothing has happened yet, but those conditions could end in a fatality or potentially significant event.
We are supporting the Benchmark Gensuite team in developing AI solutions that identify repeat events and help measure the effectiveness of the read-across process. Even if you have 10 sites instead of thousands, you still need to narrow down your view. It is really noise until you can say, “We have an issue at this site. This is happening, and we need to take action before something serious happens.”
Why is this work personally important to you?
Sandra: I have always been a futurist. I love the EHS profession and the idea of making a difference in people’s lives, health, safety and the environment. I am always thinking, “What can we do better? What can technology help us with? How can we use technology to prevent things from happening in the first place?”
When you have been involved in investigations of terrible events, you see the impact on the people involved, their families, their communities, their supervisors and the teams. No one is ever the same.
People first has really been the reason I do this. I also want to have an impact on the environment, which circles back to people and life on this planet. I do it for the people.
How would you describe the partnership between GE Vernova and Benchmark Gensuite?
Sandra: I think we are close partners. We are partners in that we are learning from each other. We can have conversations about how we improve for people, and Benchmark Gensuite is the nucleus that allows us to do that.
The team also brings perspectives from different subscribers. As a community, it is almost like an industry group of its own. Benchmark Gensuite does that for us.
It can be difficult to find the same level of entrepreneurial thinking combined with good follow-through, commitment, and depth on the bench to realize the benefits.
There are lots of quick starts and lots of ideas out there. To follow through on those ideas and have an impact, it is hard to find another company with that level of partnership.
Why does EHS expertise matter when developing AI-supported tools?
Sandra: The depth of the processes from an EHS perspective matters. These tools and applications were built by EHS people. They are unique to the EHS standards and common practices that we all try to follow.
The interconnected ecosystem of Benchmark Gensuite tools and applications is a huge strength. As the team layers on AI tools for efficiency, productivity, and risk identification, it becomes even more important for that technology to be connected and embedded.
GE Vernova has also taken the tools and customized them to our processes and our language. I do not know another tool on the market that allows you to do that.
What are you most proud of, and what advice would you give organizations beginning a similar journey?
Sandra: It has been a deeply rewarding journey. I think back to our first days in concern reporting, the work on incident measurements, and the industrial hygiene risk assessment tool. I am most proud of the work I did on the patented PSI AI Advisor and the continued development of Risk AI Advisor to become more predictive and preventive in identifying where the next severe event is likely to happen.
You do not have to be a big company to find value. You can start small, use a few of the applications, and have a big impact on your organization. I would recommend phasing it in.
Collaborative Innovation That Supports Earlier Action
Sandra’s experience shows how a subscriber relationship can continue to evolve long after implementation. Her operational knowledge helped inform the development of the PSI AI Advisor, while the Benchmark Gensuite team provided the technical resources and follow-through needed to develop and refine the capability.
The result is a more proactive way to use the information already being captured across EHS workflows. Improving descriptions, connecting data, and identifying repeated patterns can help experienced professionals focus their attention where earlier action may prevent a serious event.
For Sandra, that shared work remains grounded in a clear purpose: protecting people by finding better ways to recognize and address risk before harm occurs.
