Don’t miss out on the latest CyclingTips updates.
You sometimes hear cyclists say that their bike has become an extension of themselves. That when they’re in the zone, and everything is feeling just right, the bike disappears beneath them; that person and machine become one. But what if there was a way to integrate yourself and your bike even more closely; to control your bike with your mind?
A handful of researchers in Australia have taken a fascinating step in that very direction. Led by Josh Andres of Monash University’s Exertion Games Lab and IBM, the researchers have found a way to use a rider’s neural activity to control the motor on a custom-built e-bike. Here’s how it works.
To start with, Andres and his team took a regular Reid flat-bar commuter bike and converted it into an e-bike by installing a motor in the front hub. They dubbed the new machine “Ena”. Mounted onto the bike was an 18-volt battery and a processor unit. The processor unit in turn was connected to an electrode cap worn by the bike’s rider.
The cap used electroencephalography (EEG) to detect electrical signals in the occipital region of the rider’s brain: the region responsible for vision processing.
Human vision is composed of two parts: the central vision, where objects are sharply in focus and easily recognisable, and the peripheral vision around the edges where things aren’t in focus, but you can still perceive them.
Here’s where things get interesting: from previous studies, researchers can tell whether an individual’s vision is “peripherally open” or more tightly focused, just by reading neural activity in the occipital lobe. If the voltage is between 0.76 and 1.19 microvolts within the high alpha range of 10-12 Hertz, the individual is taking in all of their surroundings. If the reading is outside of those ranges, their attention is instead more focused.
Previous studies have shown that the “peripheral awareness state” is associated with better athletic performance, coordination and higher awareness of the environment, which has obvious implications for cycling. If you want to get through a ride safely, it’s important to be hyper-aware of the world around you and any dangers you might face.
In a conference paper about this new project, Andres and his colleagues hypothesise that changes in peripheral vision often occur due to instinctive reflexes; that when a rider goes from a peripheral awareness state to a more focused state, that demonstrates a hazard or something else of note ahead.
Imagine you’re riding along, feeling nice and relaxed, taking in the world around you, when suddenly a car turns in front of you from a sidestreet. Instinctively your attention is going to be hyper-focused on that car, and making sure you do your best to avoid a collision.
Which brings us back to Ena the e-bike. The researchers programmed the system to automatically provide motor support when the rider is in a peripheral awareness state, and to cut that support when they leave that state. That is, when the road ahead is free of obstacles and the rider is relaxed, the bike speeds up; when a potential hazard catches the rider’s attention, the motor shuts off.
And here’s the most impressive part. Because the system is tracking neural activity in real-time, it knows instantaneously whether you’re peripherally aware or not, meaning it can react and shut off the motor before you’ve even had a chance to process the threat and reach for the brakes.
As part of their study the researchers recruited 20 riders (12 male, eight female) to test the system and provide feedback on the riding experience. All participants were experienced cyclists who ride at least once a week.
Before hopping on the bike, each participant was shown a video on how to achieve peripheral awareness.
“The video invited the participant to stand up straight, fix their gaze to a point in the distance, breathing in and out slowly a few times to relax (extending their arms to the sides, and bending their hands forward to move their fingers until their peripheral view caught on to the finger movement),” the researchers explained in their paper. “Participants gradually adjusted how extended their arms were to test their peripheral vision detecting the finger movement while their gaze remained fixed in front.”
After watching the video the participants were fitted with the EEG cap then sent off to ride a couple laps of a flat and straight suburban street roughly 1.5 km long, to become familiar with the setup. Those initial laps were spent trying to achieve peripheral awareness and get the motor to kick in. Those that weren’t able to do so on those first two laps were invited to watch the video again, before trying again.
Once each rider was able to achieve peripheral awareness, they were sent out to do a minimum of six laps to become familiar with the system and the experience of riding it. Between each lap, the participants would answer a bunch of questions from the researchers.
Riding the bike
Thanks to the participants’ interview responses, we can get a good idea of what riding Ena is like. One of the most prevalent themes that emerged was the feeling that the system could react quicker than the rider could.
“There’s a minor moment of panic where you realize, ‘Hey, I need to quickly find a way to avoid this incoming thing.’ That is when the bike slows down and it gives you time to think,” said one participant. “The bike is actually responding before I’m capable of — that’s really powerful,” said another.
Participants found it took some time to get used to the experience of controlling the motor through only their mind. “You’re trying to learn how to control that part of your mind, like learning how to flex a muscle that you’re unaware of, so you [have] to try lots of different things until you start to figure it out,” said one individual.
When they were able to get it working, it was very satisfying: “It feels as all of a sudden that you’ve activated a different part of your senses, of your vision, that you didn’t know you had access to,” explained one person.
Others found it was a strange feeling: “It feels a bit surreal because you need to be in sync with your body to get the bike to accelerate,” one said, “and it then stops accelerating before I realise that is what I wanted to do.”
Interestingly, some participants remarked that they felt in-sync with the system; that controlling the e-bike was a joint effort.
“It felt like it was a combination of me, the bike and the environment,” one person noted. “I noticed when I was riding that when you are decisive, when you feel clear in your mind as to where you are going, that’s when you increase the speed.”
“I’m affecting the system,” added another, “but the system is having control over me completely because the system has more information about what’s happening than me, which makes me think the system has maybe more control over what’s happening than I do”.
So what does this project mean for cycling going forward? Well, it’s clear this technology is a long way from everyday use — if indeed it ever gets there — but it certainly appears to have some promise.
The idea of an integrated system that helps to reduce your speed before you can even react will surely have some appeal for those who have ever had a near-miss on the bike (or worse). It’s worth mentioning, though, that the Ena system doesn’t trigger the bike’s brakes — rather it cuts power to the motor. The significance of this likely depends on how experienced a rider you are.
If you’re a rider that relies solely on the motor to propel you, cutting power to the motor could well slow you down quickly enough to give you more time to react to an upcoming hazard. If you’re a more experienced rider though, and most of your speed is from your pedalling rather than from the motor, cutting the motor mightn’t affect your speed terribly much.
In this latter case, the safety benefits of the system will be more limited — the bike might well stop the motor before you get a chance to grab the brakes, but if you’re still pedalling, you’ll still be ploughing ahead.
It’s not hard to imagine a future version of this system which, in addition to shutting off the motor when a hazard is detected, also activates the brakes. Timing and modulation would be crucial though — you wouldn’t want the bike to automatically apply a large braking force when the rider isn’t ready.
One of the possible applications for a system like this would be to support paraplegic cyclists. For someone that isn’t able to pedal — and therefore activate the pedal-assist technology that’s found on regular e-bikes — the ability to control an e-bike’s motor directly via one’s brain waves seems to have some potential.
It’s worth noting, though, that while the current system does act quicker in response to dangers than the rider is capable of, the difference is only slight. If the goal is to react as quickly as possible, perhaps something akin to Auto Emergency Braking (AEB) used in motor vehicles or the threat detection used in self-driving car technology could be worth exploring. Such technology wouldn’t rely on the rider spotting the hazard, as the Ena system does.
Of course, and as noted above, applying braking forces the user isn’t expecting poses its own dangers on a bike where balance is vital for a rider to stay upright. (Perhaps an audio alarm would be a better measure when a hazard is detected, rather than automatically applying the brakes).
If you’ve got some concerns about this technology, you certainly aren’t alone. According to the researchers, several of the study participants found themselves thinking of a future “where interactive systems were able to read indirect physiological signals and automatically act on such information as our system did.
“Participants described such a future as ‘scary’ and they were wary of large technology companies misusing their indirect physiological signal readings.”
Just as peripheral awareness is mapped to a specific EEG range, perhaps other neurological states will be mapped to “specific states of our sensorial realm”. From there, it’s not hard to imagine our neural information being exploited for commercial gain or used in other more questionable ways. As the researchers note “we may need to begin defining what are our ‘inner bodily data boundaries’ [are] in order to promote our bodily data privacy.”
That’s a sobering thought, and not one that comes up when we’re talking about the average bike ride. Thankfully the founders of this current project are well aware of the issue. Hopefully such concerns will help guide future work in this space.
Regardless of how this tech ends up being used — if indeed it does — this is a fascinating project in its own right. The idea of being able to detect a rider’s neurological state and use that to support the cycling experience is both impressive and exciting, even if it’s limited in scope for now.
There’s something to be said about the way the system turns the rider’s attention inward as well. In order to use the Ena system as intended, a user must be in tune with their own body and how their body is perceiving the surrounding environment. This sense of mindfulness is something many of us love about cycling: being aware of how our body is feeling and performing, all while taking in the world around us.