Dr. Steve Wright of the University of the West of England (UWE) started his talk with a short overview of his career in aerospace involving conventional aircraft and now working with small, unmanned aerial vehicles (UAV) or drones. From being restricted to working on only a small part of an aircraft it was now possible to oversee the complete system.
Inspired by the ImechE UAS Challenge (Unmanned Aircraft Systems Challenge) a team at UWE set out to develop a simpler competition that could ultimately lead to the South-West becoming a centre of excellence in UAV technology and on the way develop teaching, commercial and military applications. Thus MAAXX, Micro Aerial Autonomous Extremes was born. The aim was to set up a competition, with as few rules as possible, in which drones would fly themselves, chariot race-style, around a conventional oval racetrack. The core rules being:
Maximum vehicle take-off weight of 3 kg.
Maximum vehicle size of 1 metre in any dimension.
No combustion-engine powered vehicles are permitted.
No human intervention is permitted once the competitor is on the course.
Unlimited attempts are permitted, at any time during the competition. Competitors may fly whenever a flying slot becomes available.
Deliberate or excessive bumping and ramming is discouraged.
Please don't be an idiot! Have fun!
The competition has been held indoors and in September 2019 is expected to be held outdoors in a large tent. The UAVs are contained within the race area by trawler netting. Three prizes have been awarded, one for the fastest lap, one for most laps flown and a 'kudos' prize voted for by the competing teams. The audience hopefully comes from schoolchildren, primary and secondary, university students and those in the industry. Competing teams can be amateurs, SME companies through to large aerospace companies.
How was this possible? Already there are many different drones available 'off-the-shelf', such as the DJI Inspire, Parrot and AR drones down to the hand-sized, £15, JJRC H2O micro-drone that he flew for us. Not only that there is a multitude of relatively cheap component systems and parts available. In a large measure these had become available as a result of advances and extensions to the capabilities of mobile phones. Some of these developments were:
32-bit processors
6-axis accelerometers
GPS
Digital Telemetry
Cameras
Brushless motors
Batteries
Open-Source Software
Of the more basic hardware, batteries rated at 8 Ah at 11.7 V, delivering peak currents of 140 A are available, as are miniature brushless motors developing high torque without the need for gearboxes. These made it possible to produce light-weight platforms capable of accelerations as high as 8.5 g coupled with rapid rates of change of acceleration or jerk, (m/s3).
Micro single board processors, such as the Arduino and Raspberry Pi, provide the computing power with add-ons giving guidance, telemetry and vision capabilities. Open-source operating systems such as Linux and FreeRTOS in a similar way provide the underlying support programming languages like C and C# , which underpin specialised drone software providing autopilot and guidance functions.
Essentially there is a large range of hardware and software modules available and with tremendous flexibility in the way that they can be put together or adapted. Because of this designing, building and operating a drone makes an ideal project as part of a degree course, where there is now more emphasis on problem solving rather than acquiring knowledge. Such a project can be used to demonstrate a student team's 'TRL 4-6'[1] capability.
It is clear that drones have great potential to perform tasks such as close-in inspection, delivery, collection and positioning of objects but there are still many challenges to be tackled, not the least ethical and legal concerns. In response to questions from the audience we were told that the aerodynamic forces were not yet the major concerns that they are in conventional airframes. The weight of the battery was the main factor determining structural strength and drag and gyroscopic forces, whilst having some effect, weren't significant in comparison. A parallel was drawn with 1919 aviation, the aircraft were capable of carrying out a useful mission but were no Spitfires!
Some of the specific problems of MAAX were discussed. Events held indoors might have no GPS coverage, bad lighting and considerable electro-magnetic interference (EMI). That meant that competitors could spend a lot of time at the beginning tuning the software to suit the environment. The circuit is provided with a red line on the floor, (which need not be followed), but one team had set up their own QR code-style ground markers. The image-processing capabilities of micro-computers such as the Raspberry Pi , (up to 70 fps), produced some very innovative solutions. Another team had successfully integrated a LIDAR, (using light pulses to measure distance), system during an event. Others used a local radio-based positioning system using the DWM1000 module.
Footnote
Technology Readiness Level:
TRL 4 Early-Stage Prototype Development
TRL 5 Late-Stage Prototype Development
TRL 6 Simulated Environment Pilot
What are Technology Readiness Levels?
This was an interesting and inspiring talk delivered by an enthusiast with almost missionary zeal. It is truly remarkable how many hardware and software building blocks are available at little or no cost. Not only that, the open-source movement, using the internet, has created a world-wide community of very talented individuals for whom 'solving the problem' is their main driver.
I contrast that to when I was an undergraduate, TTL 'chips' were the limit of affordable complexity and the largest 'flash-memory' I had was a made up of 64 diodes and 128 8 BA nuts! Oh to be that age again with today's technology!
The use of project teams in universities would concern me. An team in industry is usually headed by an experienced leader and the team chosen for their known capabilities, with some there to gain experience (succession planning?). The same can't be true at university, it just becomes a matter of chance, divas or dumbos, which makes the worst team-mates? A small 'negative' in what was a very positive presentation however.
Useful links
Competitions:
ImechE UAS Challenge Competition
MAAXX Competition
Hardware
Pixhawk – Hardware Standard for Open-source Autopilots.
JJRC H2O mini-drone review (turn the sound down!)
DWM1000 - Localino, Open-source Indoor Localization System
Software
FreeRTOS – Real-time operating system kernel
Ardupilot – Open-source autopilot software
PX4 Autopilot – Open-source Flight Control Software
A List of Flight Controller Projects
