Motorsport is a paradox. By its very nature, it pushes the boundaries of engineering yet is steeped in tradition. Ask any true motorsport enthusiast about their most memorable moments of the sport and people will be waxing lyrical about the late great Stirling Moss at Goodwood, the days when F1 wasn’t subject to so many sanctions and when racing was more about the will to succeed, not the money and sponsorship. But there are some new kids in town which some believe are hell bent on ‘destroying’ all that motorsport holds dear
by removing the ego, the personality, the life from the sport. Whether you’re a motorsport purist or like us, keen to see any level of innovation in the racing world, you’re bound to have an opinion on this apparently very contentious subject….Roborace.
All imagery is ©Roborace as far as we are aware other than as stated
Roborace: Autonomous Engineering
We’re sure you’ve already heard of Roborace, the autonomous, driverless, electric
racing series unveiled for testing during the in 2016-17 Formula E
Championship. 2019 saw Season Alpha of the championship commence and the first
race between two totally autonomous cars take place. Team Arrival, already a specialist in
autonomous vehicles and the Technical
University of Munich were the first two teams to take an identical car and
compete to win using ‘only’ algorithms and artificial intelligence. Updates on
progress for the series have been limited but the super-slick marketing
materials from Roborace HQ keep promising serious speeds and multiple sexy
looking ‘DevBot 2.0’s which can be driven by a human or ‘robotic’ driver.
As with all industries, motorsport is now heavily reliant on
technology, with software and sensors as much a part of a team’s priorities as
tyres and aero. Roborace takes that one step further and this is surely the
motorsport engineers dream? No driver ego to consider, no human error on track,
just mechanical and software engineering at its most advanced level.
You only need to look at the comments on Roborace’s You
Tube Channel and social media to see the excitement that this series is
creating. Yet opinion is still split. Is Roborace sounding the death knell of
motorsport as we know it or is it the beginning of the future? We asked our
tutors to weigh into the argument with some surprising insights.
Kieran Reeves: NMA Director of Motorsport
Studying
towards a PhD in Mechanical Engineering, focussing on energy management of hybrid
race vehicles. Former Chief Engineer and Consultant.
I’ve kept one eye looking over at Roborace as it started to develop during the 2016/17 season as the concept is the next step from my own PhD research. I have developed many algorithms that optimise trajectory and racing lines to improve lap times. These are also based on vehicle performance, energy management strategies for Hybrid Electric systems and to some extent contain an Artificial Intelligent controller as a ‘Driver’. These all reside within a Matlab/Simulink model and I can run genetic optimisation and/or particle swarm optimisation algorithms in a fully validated simulation environment to find the perfect race time. So, after that overview I return to my first comment that Roborace naturally is the next step. It uses AI systems in a real-world environment rather than just a virtual simulation.
This is what excites me, as the speed has increased in the
Robot vehicle and the first 2 car race occurred it is now starting to get very
interesting on how individual teams will build their AI systems. From a pure engineer’s
view I am looking forward to the day where we see a 20 car grid and all cars
have to adapt and utilise split second decision making. What makes a race
exciting? The drivers who push the boundaries and make the perfect last-minute
decisions that ultimately conclude with a clean but daring overtake. Developing
the AI to a place where it will behave like a human is the challenge. In
theory, with all other controllable parameters the same: car, set-up, tyres
etc. then as proposed the only differentiator is the AI system. As with my PhD
theories, it’s inevitable with time that all teams would end up with something
similar with regards the AI algorithms, therefore a perfect line and race time
is going to be relatively straight forward to optimise and for the AI to learn.
It is when 19 other cars are wanting that tarmac space that this race series
will get really interesting for the engineers.
Whether this will spawn boring races or extremely exciting
races is yet to be seen but I will continue to watch with an eager eye at how
this series develops. It is an excellent proving ground for the software
engineers and of course will ultimately help to inform the automotive industry
with regards to autonomous driving too. Exciting times ahead!
Tim Mullis: NMA Tutor specialising in Data Acquisition, Strategy & Simulation
Tim is currently a Data & Strategy Engineer working
in the European Le Mans Series as well as a tutor here at the NMA.
Roborace is great for engineers! Come on, how could it be
any better?? It’s seriously high-tech vehicles with sensors feeding information
to an AI. Real cutting-edge stuff with problems to solve. It’s a true
engineer’s dream come true!
For the engineers working with Roborace out
there, please give us more details! Where you are up to? What does your
current system actually do? What doesn’t it do at the moment? Are the
cars actually ‘thinking’? Give us answers and you would gain a great many
engineering fans. Perhaps a different fanbase to typical motorsport but we’re
with you! We want to know more about how this all works and what the impact
will be on the wider world of motorsport so please start sharing more of the
technical innovations going on.
As for ‘traditional’ motorsport, perhaps it’s
not so great. Take away the driver and the vital source of excitement is lost.
Where’s the adrenaline coming from? Surely the exhilaration of
motorsport comes from watching a human being on or in a manmade vehicle going
as fast as it will go? There’s no fear, no thrill in an autonomous racing
series. The passion to race comes from one of the most primal instincts – the
thrill of the chase, so when you remove the human element, what have you got?
Also, in many forms of motorsport, the driver
brings in the money. The sponsorship is tied to the ego, the personality.
People follow people. They get behind their favourite and spend huge amounts of
money on fandom – merchandise, tickets etc. Without that personality how much
money would be available to continue development?
However, times change and although I can’t find
a source for this, I once heard a story that text messaging was developed by
engineers to be used only by engineers to communicate and that it was felt it
would be far too ‘geeky’ for the general public and look how that turned
out! Take away the technology from motorsport and you also take away the argument
that motorsport helps push development that eventually finds its way into our
everyday lives. Environmental factors can’t be ignored – perhaps
driverless is just the first rung on the ladder to only ‘virtual’ racing
without the need to travel or use resources! Let’s face it, the current
Covid-19 situation has seen the popularity and professional standing of eSports
skyrocket.
Perhaps it is still motorsport, we just need to
get used to it as times change. Let’s not forget the first race was won with an
average speed of 15mph and in 1895 that was cutting edge and exciting!
Wayne Gater: Deputy Director of Motorsport & Big Kid
Specialist in
Aerodynamics and Electrical Engineering
Roborace has been an interesting innovation in motorsport
for me and I think it will continue to be of interest to those of the
Electronic generation. Having grown up with radio controlled cars, I was
effectively a spectator even when I had the controls so to see these cars
career around a track ‘unmanned’ doesn’t pose as a problem for me. If anything
it almost feels somewhat nostalgic, not to take away from the massive advances
in tech!
I haven’t seen the latest in developments so I’m
not sure what they are proposing now, be it pre-programmed or whether there
will be any real time remote driver involvement but I recently attended a
conference where there was a balanced amount of praise and reservation for the autonomous
racing form.
It’s interesting to think how two computers will
perform in terms of going for the gap. Does one decide on its line and the
other compromises? Which one gives? Is this a decision that the sensor make
depending on how they’re set up? Or do they both go for the line? If they do,
does one know it can’t hold its line if it runs a tyre onto the kerb or the
grass even? Depending on the crashfest potential, this may make for a winning
formula for attracting spectators! I know that when I was racing my RC10B2 (the
R/C boys will know what that is), all gaps/opportunities were taken. Whether
the car ended up on its roof or in the wall of the hall we were racing in,
scale that up and you have the recipe for some interesting watching! Although a
corner on my R/C car only cost me a week’s pocket money so there’s another
factor that may dictate how this race series would play out. As Tim said, it’s
the human element that provides the cash to compete. The tickets, the merch,
the sponsorship. Without the following will money dry up before the series has
even had a chance.
It’s incredibly interesting from an engineer’s
point of view though and I think it will certainly be a gateway for programmers
to get into Motorsport. That will in turn have a knock-on effect on motorsport
as we currently know it as all the best tech eventually filters down.
All vehicles will be autonomous at some stage so
I think this should also act as a reminder to appreciate the drive! I hate
driver aids and all the nanny systems are always switched off on my car (where
legal obvs). There’s a purity to having to control your own throttle, steering
and braking. This is something that wouldn’t impact Roborace, but if the tech
started coming into other race series, it may start to make the drivers feel
like passengers. However, minds can be changed and there is always an element
of ‘fear of something new’. I’m an optimist, I think it will be a good addition
to motorsport generally.
Roger Grimshaw: Specialist in Maths for Engineering & Fundamentals of Motorsport Technology
Grumpy Old Man and ex-Race Car Driver
As far as I am concerned Roborace is Satan's doing. I have nothing more to say.
(Editor’s note: Roger was a reasonably successful amateur race driver at one time, so this stance is unsurprising)
The Verdict
All in all, the verdict is favourable. But what
would you expect from a bunch of academically driven motorsport nuts? That’s
also why we chose them! All of our tutors, Roger included, are involved in the grassroots
of motorsport innovation and help our students to expand their learning far
beyond what is already possible. We have details of some final year projects which
explore applications of simulations far beyond what is commonplace in the industry
which we’ll soon be sharing.
The field of
engineering is in a constant state of flux and progression. It’s the true
nature of the beast. Without innovation, without improving on existing and
developing new concepts, what’s it all for? Let’s be honest, it’s still the
human brain powering the cars, just not in the form of driving. If you want an
example of engineering excellence you don’t get much more convincing than this.
According to Lucas di Grassi,
Roborace CEO and ex Formula E World Champion, Roborace poses no threat to real-world
drivers, just as SIM Racing poses no threat to motorsport engineering.
“It’s not because Deep Blue beat [Gary Kasparov] that
we don’t have chess matches any more. Humans will always compete! We are just
evolving the technology!”
We did approach Roborace for comment and we will continue
to try and get some more information about the tech and the future but for now….watch
this! The Roborace Robocar doing 175.49mph to become the
World’s Fastest Autonomous Car –
