(Image source: Kespry promotional video.)
Kespry, a Menlo Park, Calif.-based drone-based aerial intelligence platform provider, has announced new capabilities that it says significantly accelerate the assessment of roof hail and wind damage for residential, multi-family and commercial buildings. The new capabilities include on-site processing of drone-captured roof inspection data, a Virtual Test Square (VTS) to support claims decision-making in minutes, and enhanced automated hail detection, driven by machine learning. Equipped with the updated Kespry solution, insurance adjusters can now make claims settlement decisions in as little as an hour, the vendor asserts.
Historically insurers and their policyholders have been forced to rely on slow and dangerous manual assessments or earlier-generation drone inspections that can take hours to process, according to George Mathew, CEO and Chairman, Kespry. Kespry’s ability to enable insurance carriers to make claims decisions in as little as an hour and provide more accurate automated damage detection will dramatically lower the cost of claims processes and improve customer satisfaction, he asserts.
“For insurance carriers and their clients, the faster an accurate roof damage assessment can happen, the better, comments Matthew. “Touchless claims will rapidly become the industry standard.”
Jim Grabowski, a Loss Recovery Specialist with Frontline Insurance (Lake Mary, Fla.), reports that Kespry allows him to more efficiently and safely evaluate and measure a greater number of roofs on a daily basis. “I no longer have to scale ladders, chalk the roof or walk the edges of the roof pulling tapes,” Grabowski comments. “I just fly the drone. Using the Kespry drone eliminates the fatigue factor and improves our ability to professionally inspect property damage claims.”
Grabowski’s colleague at Frontline, Loss Recovery Specialist Ellen Westcoat comments, “Innovation meets roofing with Kespry. No ladders, no special boots, no measuring, while providing crisp photos and accurate measurements in a matter of minutes.”
Kespry describes key features of its new touchless inspection capabilities as follows:
Faster data processing and mobile tools deliver damage and inspection data within minutes: Claims adjusters or roof inspectors can view detailed imagery and data on the state of a roof within minutes of a 5-10-minute drone flight directly on the same iPad used to plan their autonomous Kespry flight. This critical decision-making data includes a 3-D model of the roof and high-resolution imagery that reveals any damage.
Industry-standard damage assessment tools for on-site claims decisions: Adjusters can generate virtual test squares from their roof inspection data and tag damage directly on their iPad before generating a claims-ready report. This replicates the industry-standard model of determining the extent of damage using a 100 square-foot physical sample of a hail-struck roof area. Previously, the adjuster or inspector had to climb the roof, manually mark the damage with chalk, and then apply a 10×10 inspection square. Kespry eliminates this hazardous, laborious process. Another benefit for carriers is all the data and analysis can be shared with the claimant while the adjuster is at the customer’s property, enabling them to see what the inspector sees and understand the decision-making process.
Improved accuracy for automated hail detection that further reduces time spent on damage assessment: Also announced today are improvements to Kespry’s automated hail-damage detection capabilities, driven by a new generation of machine-learning algorithms. These updated capabilities provide hail damage analysis of residential homes, multi-family dwellings and commercial buildings where a more in-depth, desk-based assessment is required. The improvements have been enabled through training data from hail damage captured in part by Kespry’s extensive insurance carrier and roof inspection customer base. The accuracy will continue to improve with customer use of the on-site VTS tagging of roof damage as it produces even more learning data for the detection algorithms.