During summer, Formula Scout was privileged enough to speak to two men at the helm of different parts of Nissan’s global research and development. Now, five months on, their exciting joint project has begun
There is a misconception that Formula E is somehow less complex, slower or easier than Formula 1, but the reality is both ask the same things of drivers in near-identical environments but just with different forms of propulsion behind them.
Increasingly, the intricacies of multi-tasking in an electric car is becoming more road-relevant than knowing how to use the power of an internal combustion engine, and as cost caps come in across top-level motorsport the smaller details are going to become more important to success a closer-fought racing future. Nissan is one of the brands that is well aware of that.
“We have been producing electrified prototypes for 70 years, and we launched the very first mass-produced very important Nissan Leaf 11 years ago, before anyone else,” Tommaso Volpe, its global motorsport director, explains to Formula Scout.
“We produced half a million cars. So we are a leader and a pioneer in electrification. Definitely for us being in FE, which is the pinnacle of motorsport for electric vehicles, is a way to promote this expertise.
“But apart from the promotional side, which is very important, it’s also a platform to develop intellectual property and research and development that can be useful for our core business.”
And this is where the possibly the most exciting motorsport-based sports science and neuroscience project ever comes in.
Before that, a brief explainer of what Nissan is already getting out of FE beyond marketing success. The development that goes into its FE powertrain, the motor, inverter and gearbox as well as the energy management software is also being taken into its automotive work, and it starts with the hardware.
“How can we make it as efficient as possible? How do we make the software for the energy management as sophisticated as possible?” hypothesises Volpe. “This is what we can transfer to the core business. How to design a gearbox, which is a level of thermal efficiency which is nearly 100% – like 97%, 98%. The same for the motor. It’s between 95-100%. We cannot disclose the number, but to give you an idea of the level of efficiency we reach. So how to make very efficient parts, this is definitely gold for us to transfer to the core business.
“The energy management software in FE, it’s very sophisticated, it generates a lot of experience that we can transfer to the normal development of our cars. And electric vehicles, when it comes to the platform, they compete on this. The successful electric vehicle platforms are the ones which have the best efficiency and the most sophisticated energy management. This is exactly the most relevant area where we could spend money in motorsport.
“Basically every single dollar that we spend is for something which is technically relevant for the core business.”
Nissan’s latest R&D step is now though going into what’s physiologically relevant to the company worldwide.
The Brain to Performance project was announced in July after months of rigorous planning with this aim: “[to use] advanced brain imaging and analysis to determine the anatomical specifics of high performance, professional drivers. The program aims to develop bespoke, optimised training to enhance the brain functions and anatomy related to driving and racing.”
The first subjects and “immediate priority” of the project are Nissan’s FE racers Sebastien Buemi and Maximilian Gunther. Buemi has been with the DAMS-run team since day one and claimed Nissan’s first win at the 2019 New York City E-Prix, while Gunther moves from Andretti Autosport for the 2021-22 season that begins in Saudi Arabia next month.
In the 1990s Michael Schumacher raised the bar on fitness for racing drivers, in the 2000s some mastered simulator tools to find an edge, and then in the 2010s an increasing number hired mental coaches to improve their abilities. For this coming decade, Nissan may be about to find the next step in performance science.
Nissan’s senior innovation researcher Dr Lucian Gheorghe is the man leading the research, having contributed to 20 previous published pieces on anatomical brain research, brain computer interfaces and the act of driving. Basically his job is to “better build the connection between people and Nissan vehicles” and that is literal.
First he pioneered the ‘Brain to Vehicle’ tech going into cars, detecting real-time brain activity and through an algorithm predicting what a driver will do next in the car, so then telling the car to prepare itself functionally for that input.
When Formula Scout got to speak with Dr Gheorghe, he said that technology was “more or less the tip of the iceberg for a long-term research project”.
It’s been a decade in the working already, and has partnered with University of Essex for the first driver tests and with EPFL, the Swiss Federal Institute of Technology in Lausanne, after Ghoerghe got his Phd in applied neuroscience there.
Brain to Vehicle focused on motor cortex activity – what happens when we voluntarily move our bodies – and where in the brain that activity was encoded. The Performance project goes beyond that.
“There’s a circuit in our brain that fires every time the same way when our expectations are not met,” Gheorghe explains. “So this second project we can use to label events in the car. Whether it’s a navigation proposing something, whether there’s a way of braking or accelerating or so on, and then you can say the driver agreed with that action or did not agree with that action.
“And then with that we could see differences in the way motor cortex prepares before doing motions. We could see difference in the way people disagree to big differences or to small differences. We could sometimes see sensitivity to very small braking actions.”
Being able to feel a wheel locking and reducing the brake pressure is crucial in racing, particularly on low-grip street circuits.
“It’s differences in ways of braking, not only where to brake,” Gheorghe adds. “And we could see some correlations with skills between these two. So we are extending now, and we are saying how can we learn from studying brain activity of very, very good drivers like the FE drivers, and this is why I am working now with Tomaso in this project. What if we can compare with these tools that we already have, the differences between average or even below-average drivers, and very, very good drivers? And if we can find some models or some specifics in these differences, how can we build an accelerated training programme in order to promote the development of the brain functions that correlate the most with laptime difference and being able to control a very good speed and so on.
“Brain to Performance is basically taking the tools that we had, developed in our long-term research project, and trying to apply it to in order to be able to develop new tools that first of all will hopefully help our FE drivers to increase their performance, and then on a longer term span that maybe at one point in time in the future will help most of our customers to drive better. Which means that we will increase the chances for our customers to be electrified, to have this access to this very special experience that is driving pleasure when being in an electrical vehicle. A long answer to a simple question.”
It’s a detailed answer, but the key for research data to be meaningful so conclusions can be reached is to have a broad and representative sample set and often repeatability. Only two FE drivers out of 100s of professional drivers (albeit of which only 25 can say they are active FE drivers) sounds very small, and there’s no way Nissan could bring in even 0.001% of the total population of ‘average drivers’ to be included in the project. Of course data protection legislation makes any scaling up very hard.
“There’s one thing to write a scientific paper, and that is exactly what you’re saying,” Gheorghe says in response to Formula Scout’s query.
“I will have to ask all of the FE teams to join the programme and to record and to write [the same us] to be able to write a paper that says ‘look, this is statistically different: these 20 guys, and these 20 guys [not from FE], their brains are this way’.
“That [scale of data gathering] happens mainly when you look only at brain activity from an electric point of view, looking from outside, an EEG [Electroencephalography electrogram reading of the brain] with an EEG device. But if you look inside, you use a device that can show you the shape of the brain, which is an fMRI [functional magnetic resonance imaging], and that shows how each brain works. Even then, it’s still rather weak in terms of publishing [a paper from this research].
“But we do have statistical tools that can show differences between one brain against another. You can find certain areas in the brain where this one is larger, it’s more developed than this other one. And for building a training methodology, this is what we are going to be using – differences in the anatomy, in the shapes of the very good drivers compared with average drivers. And then to try to build a training programme based on this shape as well, not only electrical activity.”
An fMRI detects blood flow within the brain, and therefore which areas are using more oxygen for activity on the accepted theory that the neural activity and blood flow of the brain are conjoined. To test that though, you need a huge machine which participants would have to lie motionless in for hours. Therefore preventing them from participating in many other activities…
“You can do both [activities and readings], the high-definition MRI can give the shape of each part of the brain,” Gheorghe reassures. “All the slices. And then you can use that in terms of volume analysis. And fMRI, it’s the function part which is what you’re saying with looking at blood flow and so on. An example, and this is how it leads us usually to drawing this kind of experiment, is if you ask a very good athlete to do mental training. So to repeat his protocol what he usually does before driving. He sits in the car, and he drives the lap in his mind. And what is very interesting, the better the driver, the closer the status of the brain is to real driving.”
You see this on the grid before F1 and FE races, as drivers sit down, close their eyes and go through the lap in their mind. This technique is often called visualisation, but clearly it’s triggering something similar to the usual motor cortex activity and more.
“It’s not only the brain,” Gheorghe grins. “You could see, it’s very interesting, this is another thing I’d never have the time to study here, but it’s very, very interesting. You could even see people sweating.
“So it’s not only the brain, but the whole body. The whole dynamic of the body can be put in this status that is very, very close to actual driving. And then you have very accurate drivers who can do it in real time. They would drive a 2m06s lap [in real life], we then ask them to do it in their brain, and it would be like 2m05 point something. And here I am talking about one of our Nissan GT drivers, and he said ‘I can just jump the straights’. So doing only the corners, then it’s a 1m03s. So there are ways [to do activities while prone in a MRI scanner]. And it’s the good part of working with very good drivers.
“If you ask an average driver to do that, you would see a lot of visual cortex activity. Which means that he is visually imagining himself driving. It is not really putting the brain in the status where his motor cortex moves and everything. It is only visualisation of what happens. And this is again another line of how good you are at imagining the status.”
For young drivers the benefit is clear. It’s generally accepted medically that younger brains have more neural plasticity and therefore are more likely to be ‘trainable’. Think of the saying ‘an old dog can’t learn new tricks’, then apply it neurologically.
And while professional racing drivers can do thousands of kilometres of testing, regular drivers also often use cars in a very routine way just as drivers take on the same lap again and again. Is it a comparison that would be useful for training racers?
“It could be, but what we’re looking at now with fMRI, this is a study that we have done before, asking people ‘please try to imagine to do this’ and seeing correlations with skills and so on. In this specific project with the FE drivers, we will be focusing more on brain shape, on MRI recordings. And secondly, we will be going back to electrical activity to measure the comparison with the FE drivers, we will have them driving our [research lab] driving simulator.”
Buemi and Gunther will make repeat visits to the labs to drive and be scanned, meaning it can be seen how their brains may change over time with training, if at all.
“We will use a similar process with average drivers, and then with the two drivers that we have. And then with the students we will be doing a longer term process with following up of training and following up over a period of time how they adapt to learning a new track and so on.”
There’s a fascinating scope to even turn the average drivers into racers by thought, and then see if that’s applicable on track. And with Gheorghe’s look at shape, localised specialisation could be observed in the brain where there is more white matter, which are relays which appear in greater frequency where that part of the brain is in greater use in the long run.
“It could be like that, most probably,” Gheorghe confirms. “And then the idea is how can you personalise the training to be focused on these specific areas, and how can you personalise to be focused on the areas that are not activated and should be activated. How can you train to have these optimisations from motor cortex going lower to white to grey area, then grey matter [the processing part of the brain].
“Then how can you help a person to encode all these automatic activities in a faster way than they would be doing it by simple training. Because you can train, right? If you have a person who keeps training, they will become better. But you do it over half a year. The goal is to accelerate this process. How can we help? How can we build bespoke training programmes for each of the participants in order to achieve a lot faster improvement in skill than you would do by just continuously driving?”
This is another win for racing drivers because testing time could be potentially cut at no cost to performance if optimisation can be reached faster. Just like the drivetrains, it could be Nissan is creating more efficient driving rather than faster driving.
Testing’s function in single-seaters has already changed slightly in recent years, with drivers learning tracks in simulators first and then focusing more on the cars once at a track. But using Nissan’s physiological approach is very static, whereas a lot of driver activity is gyroscopic response.
The tilting of the inner ear tells a driver if they are accelerating, braking or turning faster than the same message from their eyes can reach the brain, so surely the removal of that movement needs to be factored into the research and the fact that a racer’s muscle memory is based on the forces being applied to their body during a lap?
“Obviously if you use a static driving simulator, the development would be more about visual cortex connectivity towards motor cortex, and so on,” Gheorghe answers. “This is what’s going to happen at least this year, we’ll be focusing on this part of driving. By looking at the loop [of activity and action] from seeing, and moving. The programme is going to evolve, so we will continue working with them after the next season as well, and then let’s see how far we could go.
“Another aspect that we’re looking at is how we can use electric brain stimulation. And with this, at least you can activate areas of the brain that should be repeatedly activated. So even if the task that you do per se does not include inner ear activity and acceleration, pure physical acceleration detection, you could reinforce this loop and see what are the effects of reinforcing it with outside input. Because this is just a [simulation]. It’s how we create better networks within the brain, and if those networks can be established with stimulation, then in the end the result would be an overall better skillset.”
Nissan appears to have an answer to each question, whether it be on the scientific integrity or racing relevance, but Formula Scout’s probing goes further. Racing drivers aren’t just interacting with their car and the environment around them on track, they’re also conversing with their engineers and possibly doing so in their second or third language. Not only that, but the terms used often require additional learning and can be different from team to team.
Will that additional information processing, and how the need to multitask in the car may change the specialisation of the brain in terms of the location of white matter, be incorporated into the research too?
“That will come at a later point, and we will definitely have to be doing that [later], but [it’s separate to] trying to train in order to be better at talking with your engineer while driving, or to be better at reading the data about your battery on the dashboard while driving. Honestly that’s not straight up on the top of the list to train, but it’s on top of the list to learn.”
From a research perspective that makes sense to not add that layer of complexity, but it’s a massively important complexion in FE where drivers have to feed back to their engineers more than the other way around. And if the proposed FE feeder series that Nissan is supportive of gets off the ground, then it will be critical to find ways for drivers to learn that quickly too.
“This is very important for us [to know how they learn it]. If we can understand how they do it now, not conceptually but physiologically, what parts and how their brain works, while being able to control the car very well and do the secondary task, that is very informative for us because it’s basically a situation that emulates what happens in a daily drive.
“You are driving and you are looking at your satnav and you are talking to the person next to you and so on. At a lot faster speed, and a lot more condensed in FE, but it is the same processes. So what it means is that you have higher chances to be able to have a stronger contrast between single activity, double-tasking and so on. So all this [brain activity] that correlate with the secondary task should be more pre-eminent in a FE full driving task.”
The chain of linked events goes on when applied to an automotive setting, so much so that one of the potential outcomes of Brain to Performance could be an understanding of how racing drivers learn to co-ordinate engineer communications with their own driving then influences how future road cars introduce more complex speech and autonomous capabilities.
It’s already been applied in a similar way to Mercedes-AMG’s F1 team, where paddles and steering wheel switches have had their positions, thickness and colour optimised for driver reaction times and expectation. At Nissan, it’s been about optimising in-car displays to meet the needs of the driver’s brain. Once scanning is done, you can know how much information to show for an individual brain to process it as efficiently as possible, how dense it can be for the visual cortex but also cognitively, and when in the driving process is it best to show that info for it not to detract brain function from other parts of driving.
This level of detail can sound crazy, but it may well be a battleground in motorsport by the end of this decade if Brain to Performance can prove its worth to FE champion Buemi and his rising star team-mate Gunther, and once it’s in FE it will no doubt then trickle down into whatever junior single-seater series exist by then.
And would rivals have to invest as much R&D into this as Nissan is doing? Probably not. If “we do have a serious enough dataset in order to write a very serious paper around this area”, then Nissan will and publish it for communal gain too.
But the sharing of knowledge will now occur far more in real time rather than in academic papers published years later, and therefore there’s more data to go through to find relevant points. Both brain processing and top-level motorsport create gigabytes of data astonishingly quickly, and it’s another area where advances in the challenge of processing data could be made by Nissan if it tries to bring its access to supercomputers to both.
Potentially the software that’s developed to map the brain will then make it onto the pitwall for crunching similarly dense numbers in a live race situation. In each avenue, what may seem like an odd side-project by Nissan could actually be a masterful of way of finding extra performance. All it requires is the drivers to, almost literally, have an open mind.