Kespry Inc. today announced its Summer 2020 Aerial Intelligence updates for its drone analytics offerings, including new capabilities, tools, integrations, and third-party operator options. The Menlo Park, Calif.-based company said the release is intended to make its platform easier to deploy and use for the mining, aggregates, construction, and insurance industries.
Kespry said it has designed and sold systems integrating everything from drone data capture to industry-specific drone analytics. The company, which was founded in 2013, has customers across North America, Europe and Australia, including Colas, Grinnell Mutual, Lehigh Hanson/Heidelberg, Oldcastle, Titan America, XAP 360, and Zellstoff Celgar. Kespry raised funding in March 2019.
Kespry flies many missions
“Our initial use case in 2013 was for mining, as we gave aggregates producers a more accurate view,” said George Mathew, chairman and CEO of Kespry. “Since 2017, we’ve had full end-to-end products for enabling insurance carriers to decide whether to replace a roof, and then multifamily and commercial property assessments.”
At MODEX 2020 in March, Kespry announced that it had flown 50,000 missions for 270 companies, as well as a partnership with senseFly eBee, a fixed-wing drone for exploring large sites. During the COVID-19 pandemic, the drone analytics provider has touted its contactless roof inspections for insurance adjusters following hailstorms.
“There are early indicators that touchless inspection is becoming more and more important,” Mathew told The Robot Report. “Currently, inspectors have to be on site and are greeted by the property owner. But you could schedule outdoor inspections virtually and collect data autonomously. Summer is typically busy for property inspections, and the touchless experience has resonated quite well with property owners and insurers.”
Summer 2020 drone analytics features
Haul roads: Kespry said its drone analytics platform now makes it easier to design haul roads. It can automatically capture measurements such as road width, grade, and berm heights. The feature also monitors the road against compliance thresholds and alerts the user if an area requires attention.
DJI Phantom 4 RTK support: The Summer 2020 Aerial Intelligence revision also includes support for the DJI Phantom 4 RTK compact mapping drone. Kespry said it offers the drone as part of its standard package, along with the Aerial Intelligence platform and customer support. As part of its “Bring Your Own Drone Program,” Kespry allows companies to integrate their own DJI Phantom 4 RTK fleets with its platform.
Enhanced hail detection: Kespry said its hail-detection system now uses artificial intelligence and machine learning to help insurance adjustors make rapid and precise claims adjustments. The drone analytics system can also detect other anomalies, such as wear and nail pops.
Third-party operator support: Customers can now work with a Kespry managed fleet, their own hardware, or third-party pilot service providers, said Kespry. They can now access global pilot networks through providers such as Dronebase.
Customer training: Kespry said its seasonal drone analytics update also includes an update to its educational webinars and one-on-one support.
“Kespry is committed to continually innovating and delivering Aerial Intelligence enhancements that meet our customers’ exacting needs,” stated Mathew. “When customers work with Kespry, they’re not just benefiting from the industry’s leading platform; they also play a key role in telling us how they’d like it to iterate as they continue expanding their usage.”
Secure drone analytics lead to virtual inspections
“The person collecting data today doesn’t have to be the person doing the analysis,” noted Mathew. “We’ve seen some success with property and casualty carriers where a data collector was in the locale and the adjustment happened on the desk of a worker in a central office. Kespry’s technology is available today to do that virtually.”
What about privacy and security for drone analytics? “That was initially a concern from property owners,” Mathew responded. “We built our product with a mindset of privacy management, and the autonomous flight system includes strict geofencing. Data collection inside it was well organized so that adjudications could be done within hours, unlike the typical weeks.”
“Partly for security, no data is processed on the drone,” he added. “The data is shifted to an iPad, which serves as a secure store and forwarder to a secure cloud infrastructure. There’s a good chain of control.”
Surmounting regulatory barriers
“All drones are currently classified in the same way — a 1.2-lb. drone is treated the same as a 50-lb. drone,” Mathew said. “The safety profile of smaller drones could be treated differently. The ability to scale use cases is limited only by the regulatory landscape, including BVLOS [beyond visual line-of-sight restrictions] and certification.”
“In a world that’s increasingly software-defined, the regulations for autonomous vehicles are more generous than for our autonomous drones, which have flown 100,000 missions and covered 4.38 million acres in four or five years with a much better safety record, with no incidents,” he said.
“When we started in 2013, we even manufactured our own drone — there was no off-the-shelf technology that could meet industrial-grade parameters such as 15-mph winds,” Mathew said. “Over the past several years, we’ve shifted our focus to software. We can either provide a flight controller and data-collecting system, or customers can bring their own data, which we can process, as long as it was collected with the right parameters.”
“Drones are only part of the overall puzzle — companies still need to know the physical condition of assets using thermal imaging, video, ground or undersea robots, smartphone data, fixed CCTV, and data from drones or aircraft,” he said. “All of this data can be converged for a better understanding of status over time, which is where the value will be.”
“As the market has progressed, we’ve seen more demand and market opportunity in commercial and industrial inspections,” said Mathew. “There’s a big opportunity right now for us to expand.”
“We’re making heavy investments in volumetric analysis, and on the inspection side, it’s pretty clear that anomaly detection and object classification are still important,” he said. “For instance, after the Notre Dame Cathedral fire, there was water and fire damage, missing sections of the roof, and the need to extrapolate from the footprint of the building. It proved the value of how drones and robots can provide better insights.”