Monetizing Data from 600,000 Commercial Drones

There will be an estimated 600,000 commercial drones operating in the U.S. within the year as the result of new safety rules according to FAA Administrator Huerta. US Transportation Secretary Foxx says  “We are in one of the most dramatic periods of change in the history of transportation“.

Its common knowledge that the use cases for commercial drones is enormous. The first wave has been agricultural and infrastructure surveying and aerial photography for real estate, construction, film and entertainment industry. Use cases now coming on strong are industrial security, search & rescue and accident investigation. Not so far in the future is payload transport and finally semi and fully autonomous vehicles. Yes that driverless Uber is a drone too.

There are plenty of UAVs and sensors available to drone operators. Flight management software has matured considerably and liability Insurance premiums can now be billed prior to each flight.  However, the underlying and most significant topic for drone operators today is how to make money now that the US commercial drone door is wide open.

Lets use easy math and call it 60,000 commercial drone operators each with a fleet of 10 UAVs will begin operations over the next 12 months according to FAA estimates. Conservatively, lets also say each commercial drone operator flies only one 30 minute battery from their fleet of 10  drones per week. In round numbers today, a drone captures 10 megabyte (MB) still frame images or video each second during flight. So that’s 60,000 drone flights x 52 weeks x 30 minutes x 60 seconds x 10MB yielding a whopping 56 Petabytes (56,000,000,000,000 Bytes) per year of new drone captured raw data.  Each additional battery charge per week could generate another 56 Petabytes of raw pixel based data and still no revenue in hand for the commercial drone operator.

Most often, these large raw data sets are either processed using one of many sophisticated photogrammetry applications or the individual pixels are analyzed to find patterns or anomalies. In all cases, the software applications are highly computational in nature and require more horsepower than any typical data center server or desktop CPU can deliver. Ten years ago, Nvidia recognized the inherent CPU limitations and developed the Graphics Processing Unit(GPU) – a hardware based parallel computing architecture specifically designed to off load the high compute demands from the CPU.

Monetization in the drone sector occurs when the captured raw data is processed into a finished data product. Only at that time does the all the expense of commercial drone operations and their flights generate revenue. To support these compute hungry applications, DroneData® newest line of GPU accelerated and dedicated servers are based on Intel’s recently released i7 Broadwell-E processors with overclock CPU speeds nearly twice that of the Intel Xeon processor family which is the foundation of all cloud computing environments including Amazon Web Services.

Read more about DroneData GPU accelerated and dedicated servers with their proprietary Intel i7 Performance Boost Architecture. No upfront cost and no rental term commitments.