Session Recap Part 2: Energy Drone & Robotics Summit Asset Owner/Operations Roundtable
/At this year’s Energy Drone & Robotics Summit, we heard a panel of asset owners talk about how they use drones and the data they produce in the Energy Asset Owner/Operations Roundtable.
The panel was moderated by Suzanne Lemieux of the American Petroleum Institute (API), and the panelists were:
Shankar Nadarajah, ExxonMobil
Josh Buchanan, Chevron
Katherine Papageorge, Chevron
Marty Robinson, Dow
The content was so good that we had to split our recap into two parts.
Now onto part two where we’ll cover the panelists’ thoughts on the challenges of scaling, handling the big data coming out of inspections, and what’s next for drones and robots in energy.
EDR Summit On Demand
This summary article is great—and I’m definitely biased, but you can also watch this and other 2023 Energy Drone & Robotics Summit sessions on demand.
Challenges of Scaling
When we left off in part 1, the panelists had discussed their top use cases for drones in their organizations. Next, they talked about the challenges of scaling those high-priority applications across the enterprise.
Shankar Nadarajah of ExxonMobil jumped in immediately with how they looked ahead toward scaling from the beginning of the drone program implementation process. He said, “Right from the get go, we sat down with the right folks and we all had a candid discussion on how to do this properly and get the right architecture in place… That was what was effective in getting consistency around what our customers’ expectations are and making sure that they're aligned on the output and what they're going to receive and in turn how we're going to manage that in a secure, safe, and orderly fashion.”
Josh Buchanan of Chevron added to Shankar’s thoughts, saying, “We’re trying to do work differently, which means we're literally inventing a new wheel. We’re going back to the drawing board to say, here's the workflow I have, and then it's not just trying to inject the technology into existing workflows—to some degree it's reinventing the workflow.” He said scaling comes down to data: “At the end of the day, having effective data management and having effective AI models—it comes to solution architecture. It is so critical If you're trying to scale across the enterprise.”
Marty Robinson of Dow agreed that planning is important, but thinks there’s another key factor. “I think you also have to have flexibility because when I started this [drone program], we didn't have cloud services and we didn't have subscription models. We would not do third party, we would not do this or that. So if I built all that architecture up front, none of it would be good today. So I think there needs to be some level of flexibility, and I think it's really important that the providers understand that flexibility and become hardware and software agnostic.”
Katherine Papageorge said the challenge she has in her role at Chevron is with rethinking failure and how workers see drones. “If a trial is not 100% perfect, my operators throw it in the garbage and they say, oh, we hate that. We don't want it again. So I have to be very careful about where I deploy technology and with which groups because the behavioral attributes of trying new things, potentially failing, and then iterating—there are very few people that are willing to do that because they see it as a break into their existing work that's already bogged down and already resource constrained.”
The Data Question
Moderator Suzanne Lemieux asked the panelists how they’re managing data within their organizations and, more specifically, if artificial intelligence (AI) and machine learning (ML) are in the mix.
Katherine Papageorge said, “In general, it's about having really streamlined systems of sharing, reporting, and generating useful imagery based on that data.” She wants to make it as easy as possible, via a one-stop-shop solution, for pilots, managers, operators, and others to both use and add to the data.
Josh Buchanan pointed out that, “We've already seen ballooning data coming in full. Automation is just going to exaggerate that dramatically. So on the AI side, at least in Chevron, we're working on some use cases that are well proven across industry.” But they’re going beyond proven applications to new horizons: “The groundbreaking element for us is we have not done a lot of AI and machine learning model building on specifically unstructured data. It's one thing to do AI on structured data. It is an entirely different beast to do automated insights on unstructured data.”
Marty Robinson said he’s focused on the actual management of data. “I think that the data is important IP to each of our companies,” he said. “But at some point, the question becomes, ‘How do we keep all this? Do we keep that two hour video of that vessel inspection? Do we drop it to 15 frames a second? What do we start doing to make it manageable?’ It's starting to build up quickly, and fortunately storage is still cheap, but I see that going exponential with the more and more feature sets that these robots have.”
Shankar Nadarajah came back to flexibility. Not every customer needs the same solution for data—they don’t all need AI or automation. “We want to deliver exactly what that customer wants, almost like going to a restaurant and picking it off a menu,” he said. “I think that's where we're trying to get to is building this [data management] workflow.
What’s Ahead for Drones in 2024?
One of our favorite questions for panels is: “What’s next?” While we can’t see the future, these experts are the closest we have to predicting what’s ahead for drones.
Shankar Nadarajah said ExxonMobil has three main areas of focus:
Continuous emissions monitoring solutions specifically for methane to help with their goal of net-zero emissions by 2030 in the Permian
Drone-in-a-box and working on how to put autonomous solutions in place
Reducing demurrage costs and seeing how drones cargo delivery could help
For Katherine Papageorge, drone-in-a-box and emissions reduction are also a key priority on her team at Chevron in general. She said her main priority is, “just making sure that we have the data needed to manage our day-to-day operations.”
Josh Buchanan’s focus is on data management and much of what was discussed around that topic. “Once I have that data management, now I'm able to effectively build libraries on anomalies. So again, I know I sound like a broken record, but it's data management and it's a machine learning labeling process that we can replicate efficiently.”
Marty Robinson called ditto. “One of my two primary goals is the data management aspect, but on top of that, it's just finalizing our implementation around the world.”
Watch On Demand
Overall, the panelists emphasize the need for technology adoption, efficient data utilization, and automation to improve operational efficiency, meet environmental objectives, and enhance safety using drones in the energy industry.
All that knowledge came from just half of this great session. Check out part 1 now if you haven’t already.
You can also watch the full panel (and other Energy Drone & Robotics Summit sessions) on demand now.
