Why robots must fight real drivers on-track

There is absolutely no doubt that the biggest existential challenge facing the traditional motor industry - and, by extension, the motorsport industry - is the advent of autonomous vehicles, which are by definition designed to eliminate, or at the very least reduce, driver input.

In the past, motorsport has been threatened by external factors and crises, which it generally overcame through adaptive resilience.

The arrival of autonomous vehicles, though, takes the challenge facing the entire sport - and, by implication, its governing body, the Federation Internationale de l'Automobile, which oversees both automotive mobility and sport - to a completely different level, so disruptive will autonomous cars be.

Given that the very essence of motorsport is a driver conducting a car alone and unaided better than the competition faced in a given event, it follows that reducing or eliminating driver input reduces motorsport to, at best, a contest without sporting elements.

By the same token, slot (or radio-controlled) racing can be viewed as a sport, as can sim racing (despite the absence of automotive hardware), for in all instances driver input and skill are required.

Autonomous cars, though, fall outside the sporting sphere for precisely that reason - no driver input, being reliant purely on engineering proficiency.

Note the difference between the two activities: motorsport (for driven vehicles) and auto racing (for autonomous cars), which is best illustrated by parallels with chess.

Is chess a sport given the activity is, like motorsport, best contested in a seated position? Consider: like the FIA, the World Chess Federation is recognised by the International Olympic Committee. Indeed, chess has applied for inclusion in the 2020 Tokyo Olympics, and could also soon be recognised as a Winter Olympic sport.

As the London Chess Conference makes clear: 'The objective of a game of chess is to win. Chess involves a relentless struggle against one's opponent. There is probably no sporting activity in which two people are locked in a competitive struggle of such intensity for such a sustained period of time. One lapse of concentration, and suddenly a good position is transformed into a losing one.'

Sound familiar?

Now consider the mid-1990s Deep Blue versus Garry Kasparov chess matches: IBM programmed a supercomputer to not only take on the master, but regularly beat him. A decade later The Man versus Machine World Team Championships were staged in Bilbao, Spain, between grandmasters and computers. Both were convincingly won by machines, proving how rapidly the algorithmic technology had advanced.

Were the matches entertaining? Undoubtedly. Exhibitions? Of course. Were they, though, sport? No! But the matches did, beyond all doubt, showcase information technology on a global platform.

In 2013 Audi, working in conjunction with Stanford University's Center for Automotive Research (CARS), equipped a TT with all the kit required to scale Pikes Peak autonomously - bedecked in the colours of the Quattro S1 as breathtakingly driven up the same mountain by Walter Rohrl in 1987. AutoTT did not put a wheel wrong.

Then, as Bentley/Bugatti boss Wolfgang Durheimer - formerly Porsche R&D director, now also responsible for the Volkswagen Group's global motorsport activities - outlined during his presentation during this week's Spobis Sports Business Conference in Dusseldorf, a year later Audi staged an inch-perfect demonstration lap with Bobby, an autonomous 550bhp RS7, during a DTM meeting at Hockenheim.

The car, driven completely unaided, set a time that would frighten enthusiastic amateurs. More impressively, during wet tests at Ascari Race Resort, Bobby adjusted its corner-entry speeds according to the available adhesion and ran seasoned test drivers close.

However, due to the kit required, Bobby was overweight and lumbering. So in 2015 a successor, Robby, was developed: same power, 400kg lighter, with more sensors and speedier reactions, plus a GPS accurate to within a centimetre. The result? Robby could (and did) show some professionals the way home at Sonoma Raceway in California.

Now imagine a race for a field of autonomous RS7s (or Teslas or whatever): entertaining? But sporting? See above.

Folk who did not for a moment believe that self-parking cars would ever see the light of day doubt that autonomous racing will be staged any time soon.

Durheimer begs to differ: the 'Bentley Boy', as the German engineer is known within VW Group, believes a curtain-raiser race for autonomous cars only could be staged during the Le Mans 24 Hours weekend within five years.

ACO president Pierre Fillon does not disagree: "For me, Le Mans is a human adventure," he told reporters. "For the 24 Hours of Le Mans, we want to keep the drivers.

"But autonomous vehicles are important. The idea is maybe to have an autonomous race before the 24 hours, like a support race.

"Maybe you can imagine an autonomous car for a safety car. But for the 24 Hours of Le Mans, we want to keep the drivers."

How long, then, before Bernd Maylander is replaced by RoboBernd? Imagine the crowd reaction were a grand prix field to be led by a driverless Mercedes.

However, returning to the Le Mans support-race concept, the main issue is that staging autonomous races for identical cars is out of the question, for, without driver errors coming into play, all cars will perform absolutely identically if programmed thus. Indeed, to expect differently evokes Albert Einstein's observation of insanity: "Doing the same thing over and over, and expecting different results."

So the next step would be a Deep Blue-type race for piloted vehicles versus autonomous cars of similar specification/performance to prove who really is boss, followed thereafter by open races for purely autonomous cars to prove which brand has best mastered the tricky art (note, not sport) of algorithms.

However, adapted racetracks (or Pikes Peak) are one thing; public streets quite another. After all, what use demonstration races on ultra-smooth, beautifully cambered superstadiums with acres of paved run-off areas when autonomous cars will ply their trade mainly on cobbled roads with myriad imperfections such as drains, access covers and pocked surfaces - and slippery yellow or white lines when wet or icy?

Here Formula E, which is in the process of freeing up its technical regulations to make them more attractive to manufacturers, comes into play. The series stages its races in inner cities, on narrow closed streets - imperfections and all.

It can be no coincidence that, as the world increasingly embraces electric vehicles, no fewer than seven motor manufacturers - including Audi - have entered this nascent category to larger or lesser degrees despite it now being in its third season, nor that autonomous race series Roborace and Formula E are edging closer together.

The first question is: why manufacturers would embrace any form of autonomous racing. The answer is simple: for the same reason that IBM built Deep Blue, and the same reasoning behind all manufacturer motor-racing programmes, namely to prove proficiency.

Indeed, due to human imperfections, they do not need drivers to prove proficiency, as many manufacturers have realised. Think Abu Dhabi 2016. As for putting a face to achievements: IBM garnered greater kudos by beating Kasparov than having him endorsing software.

What manufacturers desperately need, though, is a global platform upon which to demonstrate their autonomous proficiency - think Deep Blue - and here racing, particularly on inner-city streets, perfectly fits the bill. Hence a partnership between Formula E, Roborace and autonomous racing ('auto racing', to paraphrase that Americanism) has been created, particularly pertinent given the manufacturer presence in FE.

"Given current perceptions of autonomy, would any responsible parents send their three children, aged, say, 15, 12 and eight, off to visit Granny living 50 miles away in an autonomous car?" asked Durheimer rhetorically at Spobis.

"Now, consider watching 20 cars contest an autonomous race on Sunday, one held in total safety, without accident or incident. Would parents pack their kids off to Granny under such circumstances? I believe 'yes' is the answer."

However, such campaigns cost huge piles of money - and, where in the past manufacturers competed only against each other, now they have all automotive brands to contend with, plus various Silicon Valley upstarts such as Apple, Google, Tesla and Uber, all of whom are developing autonomous technology.

Worse (for car companies) is that five of the USA's top 10 cash-on-hand hoarders in 2015 were comptechs with holdings totalling half a trillion dollars. No car company even got a look-in.

Car manufacturers will therefore turn to the one arena they know and understand - racing - as a platform, and can be expected to divert existing motorsport budgets to 'auto(nomous) racing', as much as seven have already allocated budgets to Formula E.

That does not bode at all well for motorsport overall, only for its autonomous versions.

The second question is, then: who would watch autonomous racing, to wit 20 driverless cars careering around an inner-city street circuit?

Pose that question to the 170,000 visitors and 7000 media members who attended the Consumer Electronics Show held over three days in Las Vegas last month, or the similar number expected to visit Shanghai's CES in June.

That these shows are staged in the world's two largest car markets is not coincidental, for, as Forbes magazine reports, "In less than a decade, [CES] has rapidly become one of the most important shows in the world for demonstrating future transportation and mobility technologies."

As though proof were required of CES's place in the automotive landscape, consider that Renault/Nissan/Mitsubishi CEO Carlos Ghosn and Ford boss Mark Fields keynoted at CES, with the former contending that "over the past few years CES has started looking more and more like a motor show", before introducing "the disruptive triangle of autonomous driving, electric vehicle and connected cars".

During his presentation, Ghosn announced partnerships with NASA and Microsoft to develop 'Seamless Autonomous Mobility', which combines in-vehicle artificial intelligence with human support to help autonomous vehicles make decisions in unpredictable situations, and expand the database of in-vehicle AI.

Talking in Dusseldorf, Opel's chief marketing officer Tina Muller confirmed that parent General Motors would shortly begin testing its autonomous technology, but stressed that shared rides were only economically viable if such providers removed the biggest cost driver - the human driver.

With Deep Blue, Big Blue and NASA involved, and Uber being the talk in car-company boardrooms, can you see where the industry is headed; how does Future Formula 1 even fit into this landscape? Indeed, will inner-city Future Kid, with a craving for consuming media via wearables and a predilection for shared rides as a direct result of having zero aspiration for car ownership, even care about motorsport, including F1?

There will, of course, always be motorsport enthusiasts, just as music aficionados prefer the richer sounds of vinyl to the convenience of streaming, just as horse lovers prefer the gallop of hooves to the vibrations of cars, and so motorsport will continue in some form, followed by hard-core petrolheads.

But, as Durheimer observed: today there are no work horses, only sports horses.

Motorsport will, though, face increasingly serious economic, political and social challenges from 'auto racing' events contested by major motor manufacturers (including Tesla), plus the likes of Apple, Google, Microsoft and Uber, and risks being displaced in the audience/media stakes much as motorsport eased out equine sports at the turn of the 19th century, when motorsport as we know and love it emerged.

Tyre companies also face challenges: do they continue developing degrading rubber to appease fans, or invest in sensor-equipped 'smart' tyres that 'read' adhesion levels, then feed such data to a battery of black boxes. And, who to sponsor - F1 or A(utonomous)-GP?

The FIA, too, will need to adapt. Speaking to Autosport during the body's annual sport conference last June, president Jean Todt dismissed any threats from autonomous cars, saying, "[Motorsport] is [about] drivers, it's competitors. Our members on the sport side have a racing licence, so we don't give a racing licence to a robot."

That said, given that the last word in the FIA's full name is 'automobile' and that global mobility and resultant road safety make up one of its three pillars, world motoring's governing body is perfectly placed to sanction both motorsport and autonomous racing.

Whatever, it is clear which way motorsport's winds are blowing. Earlier this week my colleague Jonathan Noble argued that "Formula 1 must become road irrelevant". Based on the foregoing it seems it has managed that already.