Drone racing looks really cool, particular the First Person View (FPV) stuff - it’s Star Wars made small and real. Not convinced? Check this out:
I’m a Senior Consultant in our Wireless and Digital Services division – a wireless guy – so my mind immediately goes to that First Person View and I think about how the video is transmitted from the drones to the pilots. Right now these transmissions are typically sent using analogue A/V, which tends to be quite reliable and high quality, but suffers from low capacity. That’s why in most cases a single race can only support eight pilots. This is an obvious area where a digital radio system could significantly improve the capacity.
A quick 'back of the envelope' calculation suggests to me that a digital radio system with a modest spectral efficiency of 1bit/s/Hz would be able to support 160 simultaneous 2Mbps data streams in the 5.8GHz band that’s currently used for drone racing. That’s 160 drones racing in the same warehouse. Now you’re talking! Even allowing for the shared nature of this spectrum and various other efficiencies, it’s clear that modern digital radio systems would ensure that the video transmission is not the limiting factor in the number of drones in a race.
One of the main advantages of a digital system is that multiple streams of information can easily be multiplexed over the same channel. This opens up the potential to improve the user experience by augmenting the video stream with more information, for example real-time position and velocity information.
So, if we wanted to make the FPV digital, and massively increase the number of simultaneous racers, this digital radio system would need to have the following attributes:
- Low latency - for typical real-time control applications the round trip time must be less than 1ms
- High reliability - it may not be life threatening but losing control of the drone because you lose the video would be monumentally annoying (and would be heavy on the wallet too)
- High throughput - each video stream will need a few mbps for good quality video
- High capacity - the more racers there are, the more total capacity would be needed in the system
What’s striking about these attributes is how closely they align with the requirements for next generation 5G systems, in particular the low latency and high reliability requirements. These latter two requirements are impossible to meet using current off-the-shelf ISM band technologies, such as Wi-Fi, and even the latest cellular technology (LTE) would not provide the required latency. Although the mobile industry has defined requirements for 5G latency and reliability, at this stage most of the progress towards next generation systems has been focused on throughput and capacity so there is no clear way forward to meeting these requirements. In any case, 5G radio systems are still a number of years away from being deployed.
What makes this particular scenario resonate for me (beyond the fun films) is that we’re often approached by clients with a set of rather specific needs for a wireless radio system and who can't get an existing off-the-shelf solution that will meet all of them. One recent example was Ocado, the world’s largest online-only grocer. Ocado’s warehouse automation solution uses autonomous robots moving around above a grid, storing and retrieving crates stacked within that grid. Ocado needed a radio control system to coordinate thousands of these fast-moving machines to within a fraction of a second. We created a custom solution that was a breakthrough in radio design – the most densely packed mobile network in the world.
Projects like the warehouse automation at Ocado tell me that FPV drone racing could leap forward in the coming years, with scores of drones racing simultaneously. I may have to invest in a little TBS Vendetta…