AiThority Interview Series With Tomer Kashi, Co-founder & CEO, SkyWatch

How did the idea for SkyWatch come about?

Around the time I got my FAA certification, we started talking to a lot of drone operators and felt that there’s a pain, mostly around the lack of education for safety among pilots, which led to increased regulation.

We felt that our expertise in complex algorithms and machine learning can help alleviate that pain —so we set about to build what we later recognized as the first iteration of SkyWatch.

Who do you see using SkyWatch? Can recreational and commercial pilots both find use?

We see SkyWatch as a platform for all drone flyers. It doesn’t matter if you’re a recreational pilot flying for the first time, or a commercial operator doing a gig for a client – SkyWatch can benefit both!

Our proprietary platform can help all pilots plan better for the flights, get real-time hazard warnings, and see post-flight risk analysis that can assist them to become better, more aware pilots.

How does SkyWatch leverage AI to map the risks for drone users?

Great question! First things first, we are using machine learning to create the risk model itself, meaning we insert all of the predictive sensors into our model together with a crash/no crash indication, and then, by performing “feature extraction”, we manage to find out what are the major factors that influence the risk.

For example, we found out flights that are done for more than one time in the same location are two times safer than flights that are done in a new location for the first time. Or, we showed that when the drone flies far away from the operator there is much more chance for it to crash, and we know how to quantify that.

Another thing that is in the making is our proprietary crowd heat map, we use artificial intelligence in order to predict how crowded a specific area is going to be in a specific time and we’ll be able to assess the risk and price the policies accordingly.

What kind of information does the app provide the pilot, pre and post flight?

The info is divided to 3 parts; pre-flight, during flight, and post-flight. In pre-flight mode, the app lets an operator select his flight zone, and shows them on the map what hazards they can expect – in terms of roads with heavy traffic, no-flight zones, airports, schools, hospitals etc.

During flight, we monitor the behavior of the operator and the drone itself, comparing the flight path to the planned area to see if they’re flying out of bounds, warn them when they fly above roads or when they fly in an unsafe way (for example, zig-zagging rapidly).

Post-flight, we combine all factors, including battery life and other statistical data pulled from the drone to create a unique “Safety Score” which we use to rank the flight.

You take into consideration the skill of the pilot. How is this calculated, and what is the impact on cost of insurance?

After a pilot accumulates several safe flights (with a high enough safety score), we can offer the pilot an automatic premium rate discount for being a safe pilot!

We calculate this safety score based on multiple factors during the flight. For example, factors include not flying above crowded areas or heavily congested roads, landing with high enough battery, keeping the drone in a close sight etc.

What factors does SkyWatch consider when creating a flight plan?

When an operator plans a flight, we take into account several factors, such as the area he’s flying in, if there are any aerial hazards around them, any major populated areas, such as schools or hospitals, or congested roads.

We combine all of these factors to show the operator the potential risk of flying in that area, as well as the cost of insurance.

How do you see the drone industry growing in the next five years? How will risk assessment technology impact this growth?

We believe that drones are going to become even more common and frequently used. You can already see drone service providers that get beyond visual line of sight and autonomous flight waivers, and Goldman Sachs recently estimated the industry will grow by $100B in the next 5 years.

As more drones are going to be in use, the importance of safety and being able to predict and avoid risk is going to be crucial to growing the entire industry.

We hope that our platform and safety score can help solve this growing pain, by creating a new benchmark for safety.

In which geographies are you seeing maximum traction? Are there specific industries that you are focusing on?

We’re now in Beta stage, collecting data from operators across the globe, improving usability and ironing the final kinks. We aim to launch the app officially later in Q1. Initially we’ll be offering insurance in the US, and hope to expand to other territories soon.

In terms of industries, we believe our platform can assist commercial operators from all industries, but we expect to see more traction coming from photographers, inspection and surveying services.

Congratulations of landing Seed Funding! What’s next for SkyWatch?

Thanks! We still have a long road ahead of us. Our next milestone is announcing a partnership with one of the major US insurance companies, to allow us to offer usage-based, on-demand, easy-to-purchase insurance that’s priced for the operators’ needs.

The product will take into account all of the safety indicators we mentioned before, and will allow pilots to pay not only “when they fly” – but “HOW they fly” – which is something we’re very excited about.

After that, we will be looking to enhance our platform capabilities in order to offer more and more services for drone service providers, ensuring “we’ve got you covered” for all needs, not only safety/insurance.

Thank you Tomer! That was fun and hope to see you back on AIthority soon.