AI’s Real Role in Commercial Drone Operations: Handling Data at Scale

Commercial UAV Expo panelists point to data analysis, scale and human oversight as the near-term value of artificial intelligence.

As commercial drone programs scale, artificial intelligence may be solving a problem that drones themselves helped create: too much data for people to reasonably review.

That was one of the key takeaways from a session at Commercial UAV Expo in Las Vegas last week. The panel, “AI and the Commercial Drone Industry: What it is Delivering Today and Where it Goes Next,” focused on the gap between AI’s promise and its practical use in commercial drone operations.

For operators, one of the clearest applications is already here.

“AI is great to analyze the data,” said Ismar Avdic of Stratus Autonomous. “With drones, we have so much data – no one has time to analyze 6,000 pictures. That’s what AI can do.”

More Autonomy Means More Data

The challenge becomes more important as drone operations move from individual missions toward persistent and increasingly autonomous operations.

Vikhyat Chaudhry of Buzz Solutions described utility customers using drone-in-a-dock systems at substations. These systems can conduct frequent inspections and transmit data to the cloud. As operations scale from one operator managing one aircraft toward one person overseeing many drones, the amount of data collected grows quickly.

Chaudhry said more frequent inspections can generate 10 or 20 times more data. That creates a second scaling challenge: collecting more information has limited value if organizations cannot efficiently analyze it.

AI can help close that gap.

Avdic pointed to processing large volumes of data as one of AI’s strengths. Chaudhry said AI can recognize patterns and help standardize processes. In infrastructure inspections, those capabilities can help identify defects and determine which findings require attention.

Newer AI models can also add context. Chaudhry said Buzz continues to rely on deterministic models for mission-critical work, where there is little room for error. Generative AI can augment those systems, helping customers understand whether a finding represents a lower-priority problem or requires engineers to respond quickly.

Keeping the Human in the Loop

The panelists were also consistent about the limits of AI.

For mission-critical applications, Chaudhry said organizations need to be able to explain and justify conclusions. The result cannot simply come from a “black box.”

Desiree Eckstein of On the Go Video emphasized the continuing role of subject matter experts. Workers need to understand the data and verify that the information coming from AI systems is correct.

Chaudhry agreed. While AI models may eventually become more widely available, he said subject matter expertise remains essential to guide their use and connect the technology to business outcomes.

That suggests that AI’s near-term impact on commercial drone operations may be less about replacing people than changing where their time and expertise are used.

Avdic compared the shift to earlier automation in aviation. Autopilot did not eliminate pilots, he said. Instead, it changed their role.

The same pattern may be emerging in commercial drone operations. As autonomous systems make it possible to inspect assets more often and at greater scale, organizations can quickly find themselves with thousands of images and other data points to review.

AI can process that volume, recognize patterns and help identify what deserves attention. Human experts can then focus less on searching through thousands of images and more on deciding what the findings mean and what to do next.

For an industry working to scale drone operations, that may be AI’s most immediate value.