Drones, AI and Wind Power: An AC Interview with Yan Ke
When you think about dams, bridges, railroads, or power lines; it's an expensive effort in resources and time to examine these massive structures for defects or damage. But it's even more expensive to find out once they fail. So physical inspection is one of those vitally necessary, complex, pain-in-the-ass business processes that every infrastructure organization must do.
Enter drones, computer vision and hard core AI in the cloud.
I met Yan Ke at the Seattle BizTech Summit earlier this month. We spoke about AI, computer vision, drones and wind turbines. Yan is the chief technology officer (CTO) of a Shanghai & Seattle company called Clobotics (cloud + robotics). Clobotics specializes in Drone-Data-as-a-Service for large infrastructure inspections. It's first major industry sector is wind energy.
Specifically, Clobotics drones inspect the huge 80-100 meter wind turbines in a wind farm. It typically takes a human operator about a day to climb a tower and inspect the blades for cracks, sun damage and other factors that might impact their performance. The drone can do it in roughly 20-30min.
To do this, Clobotics merged flight algorithms with computer vision. After mapping the shape of the turbine, the drone hovers about ten meters away as the camera shoots images that are scanned for defects/damage by neural nets in the cloud. The output for the infrastructure owner is a detailed 3D map of a turbine with all of the various dings, rips and scratches documented. From that point, the owner/operator can schedule repair or maintenance much better.
Yan and I spoke in greater detail about computer vision, China's huge investment in wind power and renewable energy, plus how to apply drones and computer as a general business capability. Enjoy.