By Ofer HaCohen, Director, AT&T Foundry in Israel
In 2017, the Federal Aviation Administration (FAA) reported a monthly average of 250 safety incidents where drones flew too close to manned aircraft, a few of which caused operations at major international airports to come to a screeching halt.
Israeli startup Vorpal specializes in tracking and, ideally, preventing those near misses. Their drone detection and geolocation tracking solution VigilAir uses a geographically distributed network of sensors that scan relevant frequencies to identify drone transmissions. This allows them to identify and track drones and their operators in near-real time.
VigilAir has already been useful to commercial drone monitoring, airports and public safety law enforcement agencies across the globe. But the FAA expects the number of drones in our airspace today to increase as much as threefold by 2023 as commercial drone operations become more and more prevalent.
To prepare their solution for the future, Vorpal is collaborating with AT&T and Microsoft to test how edge computing could equip them the ability to track thousands of drones at any given time.
The answer’s in the clouds
Today, the ability to pinpoint the location of drones in near-real time is key to VigilAir’s success. To ensure VigilAir can seamlessly maintain those same capabilities while monitoring thousands of drones, they’ll need robust, low-latency and high-throughput network capabilities right on the device, at the intelligent edge. So Vorpal is looking to the clouds with AT&T and Microsoft.
Each of Vorpal’s sensors is equipped with computing hardware that processes their location-tracking software. The more drones in the sky, the more compute power needed to handle all that data.
What if we could pull much of that compute out of the sensors and put it in the network? Simplify the sensors, reduce their cost, and transform that stranded processing power into a shared resource.
That’s where network edge compute (NEC) comes in. Together with Microsoft, the AT&T Foundry is testing how to bring NEC capabilities into AT&T’s network with Microsoft Azure’s IoT and AI services and Azure Stack hybrid technology. By deploying Microsoft’s advanced cloud services closer to the edge of the network – in this case, within the geographic locations where customers like Vorpal need them – NEC could allow businesses to access low-latency network compute at a fraction of the cost of traditional, embedded processing.
What have we learned?
In collaboration with Amdocs, our team at the AT&T Foundry in Israel kicked off our testing with Vorpal by examining various performance indicators of their VigilAir solution in the Microsoft Azure public cloud.
Using four drones, we then tested various performance indicators over Azure Stack in a NEC environment at our Foundry in Plano, TX. Vorpal deployed their VigilAir sensors over our LTE network across one square mile around our building. Through this NEC environment, we experienced a 40-50% network improvement over testing in the public cloud, with speed tests showing 24-40 milliseconds of latency. But our testing also showed low variance in latency, which is critical as it ensures Vorpal’s ability to provide a consistent quality of service.
Where edge, 5G and drones intersect
Our ultimate goal is to help Vorpal scale their solution to provide the most business value, now and in the future.
Once 5G networks become ubiquitous, their even faster speeds and lower latency rates will only amplify NEC’s capabilities. This 5G and edge combo means Vorpal could deliver the near-real time situational awareness and tracking capabilities for thousands of drones, reduce the cost of multiple sensors and large-scale deployments, and unlock new solutions for commercial drone use.