“Where it gets really exciting is where it combines with machine learning and artificial intelligence to open the potential for predictive maintenance on projects. Eventually, no human intervention should be required at all to assess or even navigate hazards. A drone will be able to do it on its own.”
Before that can happen, drones must become big data enablers. The challenge facing the drone industry centers around transitioning away from short-lived photography, to generating increasingly structured spatial data about the world around us. Drones are essentially data collection vehicles, combining near infinite mobility with a budding sensor marketplace. By autonomously capturing the same sets of data over time with precision and accuracy, the opportunities of big data can be applied to physical world information - increasingly requiring computer analysis to reveal changes, patterns and trends.
“By tackling autonomous mission planning and automating the data supply chain, we’re unlocking structured spatial data at scale,” says Brown. “We’re exposing millions of data points to machine learning algorithms that detect change, recognize patterns and unleash the power of AI.”
“It’s their ability to get up close and to identify critical information not previously detectable that makes drones so valuable across so many industries, reducing costs, increasing productivity and operational efficiency, all while keeping workers safe,” he concludes. The combination of big data, digital transformation, unlimited processing power, edge computing, vision-based machine learning, and autonomous mobility will enable robotics to augment and replace mundane human activities — leading to an exponential increase in business intelligence, productivity, insight and societal value.
The uses will broaden further. While human intelligence is required to derive insights from the visual data gathered by drones, Hangar has identified what it calls 4D Data, in which computers cue humans to act. The next frontier would be autonomous robotics, which collect, understand and react to visual insights in real time, without the involvement of humans.