Thursday 13th September 2018

BMW's autonomous R1200GS on video

You thought autonomous cars were odd? Wait until you've seen the bike that does the same...

---

Last year BMW revealed to the world the R1200RS ConnectedRide prototype; with a host of technologies including Vehicle to Vehicle Communication and a Differential Global Navigation Satellite System, it aimed to make riding a motorcycle safer, more connected and less dangerous. A year later and that concept has taken another leap forward - it can move autonomously.

Yep, here's a bike that doesn't require a rider. Shown at BMW's Motorrad Tech Day in Miramas, the BMW - that's now an R1200GS - can start, accelerate, corner and stop independently. As you can see in the video below, it's equal parts impressive and spooky. Interestingly BMW says it "is by no means aiming for a completely independent motorbike." Instead, as with the original concept, the ambition is to make cycling "more comfortable and increase the riding pleasure." It will be tested to discover more about dangerous situations and the appropriate safety tech to support the ride in those.

There isn't much more issued in the press material, though further details are discussed in the video. Without plans to fully automate motorcycling, surely the introduction of new safety measures, improved stability and more control is a good thing? Over to the PHers on two wheels...

 

[Onehp]

Current tech is great.

Biggest danger remains the frontal lobe brain cells allowing for too much right hand twist.

Reminds me of my radiocontrolled motorcycle toys 30years ago that you had to control by weight tranfer of the battery cradle...

[Mr_Sukebe]

Good showcase of capability.

I see the biggest use being for city scooters and similar.
Uber are already looking a renting electric scooters in the US. Give the scooter this technology and you can have your scooter drive to meet you, then after you've finished with it, it wanders off to a holding area where it can get re-charged before going on to the next bunch of customers.

So probably a bit pointless on the bike in question, but there's a lot of VERY serious applications for the capability.

[horsemeatscandal]

Safer AND less dangerous? Sign me up.

Interesting technology, monstrosity of a bike.

[black-k1]

Definitely a solution looking for a problem.

Although …

for all those who buy bikes, put them in the shed/living room and don't use them this could be a way of getting miles on without having to actually ride the bike!!!! No more embarrassment at the lack of miles when the MoT is renewed.

[sidesauce]

black-k1 said:

Definitely a solution looking for a problem.

How so? It's clearly stated by Stefan that their goal was not to develop a fully automated bike but to improve safety. Given the amount of motorcycle injuries and fatalies due to human error I fail to see how this train of thought is a solution looking for a problem; rather, it's an attempt at finding a solution to a problem!

Stuff like this is fascinating as given it's early days in development the tech will improve extremely quickly over the next few years. The last line in the commentary "and self-driving motorcycles on the road, remain, of course, science fiction" should really have finished with "for now...".

[xjay1337]

Cool.

I remember watching one race Rossi or something and it was slower but not as bad as you'd have thought.

[Andy XRV]

I do hope they add a "filtering mode" which either makes the bike nip through traffic properly or if not, when there's a bike behind it gets out the bloody way.

[black-k1]

sidesauce said:

black-k1 said:

Definitely a solution looking for a problem.

How so? It's clearly stated by Stefan that their goal was not to develop a fully automated bike but to improve safety. Given the amount of motorcycle injuries and fatalies due to human error I fail to see how this train of thought is a solution looking for a problem; rather, it's an attempt at finding a solution to a problem!

Stuff like this is fascinating as given it's early days in development the tech will improve extremely quickly over the next few years. The last line in the commentary "and self-driving motorcycles on the road, remain, of course, science fiction" should really have finished with "for now...".

I would suggest that most motorcycles are bought as "toys" by those who want the enjoyment of riding their bikes. As such, I would think there will likely be little to know interest in a self riding bike other than for the novelty value. You only have to look at the resistance there has been from the general motorcycling community to ABS and traction control to understand that electronic "riding aids" have an up hill struggle even when they enhance rather than detract from the riding experience.

[The Selfish Gene]

not in a million years would I allow that anywhere near me.

Horrendous technology in cars, but on a motorbike - the most pointless technology in the world.

Nobody will use it. Well maybe those people that like jumping of mountains dressed as squirrels.

They seem to have a death wish.

[phil4]

Unless I've misunderstood this is a long way behind the curve.

They modded RC cars to follow a path round a race circuit years ago. Granted this has to lean, but it's really no more than that. And as someone already said Yamaha's Motobot was racing Rossi last year: https://www.youtube.com/watch?v=BjZPvXKewFk

I'm not saying it's a doddle, but at a high level, all the bike needs to do is check it's GPS vs the plan and move itself back to that planned route when moving off it. The hard bit vs car is getting the lean right, but that's not impossible, clearly.

As autonomous cars have shown - there's a heck of a lot more to it than that, and usually that heck of a lot required lots of cameras, some lidar etc. We're still some way off cars managing this safely and in all scenarious, let alone somehow squeezing down the sensor and computing power to fit on a bike, and adjusting to the challenges it has.

And that's before you consider a good many people wouldn't want to use it anyway.

Honda seem to be more onto something with their ride assist: https://www.youtube.com/watch?v=VH60-R8MOKo So it doesn't fall over and once you're off it, you can guide it into a parking space/garage. Just the not falling over would do.

[Prof Prolapse]

Is "Stefan Hans" not the most German person you have ever seen? I mean, the only way he could look more German is if he he did that interview with half a plate of CurryWorst and David HasselHoff playing in the background...

That observation aside, looks fascinating. The physics of bikes is obviously a lot more complicated than car, so although not new but still very impressive to me. I'd be curious how it will filter into rider aids. A "I'm-fking-knackered-just-make-sure-I-get-home" button on a GS would be great as a bit of reassurance against rider fatigue.

[Prof Prolapse]

phil4 said:

Unless I've misunderstood this is a long way behind the curve.

They modded RC cars to follow a path round a race circuit years ago. Granted this has to lean, but it's really no more than that. And as someone already said Yamaha's Motobot was racing Rossi last year: https://www.youtube.com/watch?v=BjZPvXKewFk

I'm not saying it's a doddle, but at a high level, all the bike needs to do is check it's GPS vs the plan and move itself back to that planned route when moving off it. The hard bit vs car is getting the lean right, but that's not impossible, clearly.

It's not about cutting edge technology riding a bike, it's about using existing technology and fitting it to a road going motorcyle to explore the feasibility of using it in rider's aids.

Also, "the modded RC cars" therefore it's easy although "granted this has to lean"? What?

I mean it seems bloody obviously even for a grade C student of physics like myself that the engineering required to move a inherently unstable 350kg, two wheeled, object, within an exisiting framework, is vastly larger than the requirements of moving a 1kg, stable object at about, what, half the speed?

It's not earth-shattering developments, but let's not belittle the achievement here.

[phil4]

Prof Prolapse said:

Also, "the modded RC cars" therefore it's easy although "granted this has to lean"? What?

Basically what you're saying, that the bike isn't stable and needs to lean through the bends. I'm talking about cars like this: roborace: https://www.youtube.com/watch?v=QtVbch-02Fs

Prof Prolapse said:

I mean it seems bloody obviously even for a grade C student of physics like myself that the engineering required to move a inherently unstable 350kg, two wheeled, object, within an exisiting framework, is vastly larger than the requirements of moving a 1kg, stable object at about, what, half the speed?

Maybe by mentioning RC cars I've confused things... I meant RC fitted to full sized cars, then driven by remote control. How that's now evolved (as per roborace above) into "self driving cars" similar to this bike... ie. following a pre-set course... so it's not really comparing 1Kg with 350Kg. Also, perhaps you're being a bit misleading with the unstable... bikes seem pretty stable once moving.

Anyway, I hope that clears up the confusion, apols for that.

[Prof Prolapse]

I’m not confused, I just think you’re an ignoramous.

[Biker's Nemesis]

Ideal for the vast majority of fat useless wkers that ride them.

Now where's that R1 that rides itself.

[big_rob_sydney]

black-k1 said:

I would suggest that most motorcycles are bought as "toys" by those who want the enjoyment of riding their bikes. As such, I would think there will likely be little to know interest in a self riding bike other than for the novelty value. You only have to look at the resistance there has been from the general motorcycling community to ABS and traction control to understand that electronic "riding aids" have an up hill struggle even when they enhance rather than detract from the riding experience.

Hmm...I know many people who ride bikes for a living (eg police motorcyclists), and I'd suggest that there would be interest at some points where fatigue may cut in to enable safe continuation. The alternative would be on long distances to be forced to stop. In that sense, this enables more riding.

But as far as resistance to other rider aids, I couldn't disagree with you more. I'd been a huge fan of bikes like the older generation R1, and recently moved onto a modified BMW S1000RR, which has a lot of the bells and whistles. I can tell you for a fact that I will never go back to a bike without the electronics on it again, for a whole host of reasons, and this is after 30+ years of riding (I'm 48), including stints racing around Eastern Creek in Sydney on superbikes.

[black-k1]

big_rob_sydney said:

black-k1 said:

I would suggest that most motorcycles are bought as "toys" by those who want the enjoyment of riding their bikes. As such, I would think there will likely be little to know interest in a self riding bike other than for the novelty value. You only have to look at the resistance there has been from the general motorcycling community to ABS and traction control to understand that electronic "riding aids" have an up hill struggle even when they enhance rather than detract from the riding experience.

Hmm...I know many people who ride bikes for a living (eg police motorcyclists), and I'd suggest that there would be interest at some points where fatigue may cut in to enable safe continuation. The alternative would be on long distances to be forced to stop. In that sense, this enables more riding.

But as far as resistance to other rider aids, I couldn't disagree with you more. I'd been a huge fan of bikes like the older generation R1, and recently moved onto a modified BMW S1000RR, which has a lot of the bells and whistles. I can tell you for a fact that I will never go back to a bike without the electronics on it again, for a whole host of reasons, and this is after 30+ years of riding (I'm 48), including stints racing around Eastern Creek in Sydney on superbikes.

I think that bikes used by police, paramedic and other "professional" motorcyclists represent a very small percentage of total motorcycle sales., even among larger capacity bikes. I think they'd also be the ones least "happy" with handing control of their lives over to any "active" automated process because their levels of training are likely to make then significantly better at managing risk than any technology we have now or a are likely to see in the foreseeable future.

As for electronic aids, you may love them but most motorcyclist don't. I am an electronic aids fan and have ABS bikes since the mid 80's when they fist appeared. However, despite 30+ years of evidence to show the advantages of ABS on bikes, it only became mainstream when the EU mandated it's fitting to new bikes. When there was an option, non-ABS equipped variants often out sold their ABS equipped equivalents even when the additional cost of the ABS was minimal, and that was bikes aimed at the motorcycling demographic deemed most likely to accept ABS.

Likewise with traction control. I would suggest that, anecdotally, one of the most asked traction control questions of forums (including this one) is "how to I switch my traction control off?". That is despite hard evidence of even god-like experienced motorcyclists ending up on their backside in situations where traction control would have unquestionably saved them.

[Biker's Nemesis]

I've only asked how to turn wheelie control off. The nearest to having an accident on my MT 10 was caused by ABS at very slow speed where it activated causing my to run on over a junction.

[sjtscott]

Urghhh can't think of anything worse than a bike that can ride itself - its completely pointless. The fact its a GS for me makes it even less appealing. For once the panniers are actually being used for something needed/useful rather than the packed lunch.. the crap it needs to ride itself lol.
I'd be happy if they added a 'get the f*ck out of the way cos you're holding all the other narrower bikes up' filtering mode for city use as an option that would be good Or a 'nope you're not gonna fit through that gap in a million year but hey look there's one here you've completely missed that your tank will fit through' mode too LOL

[black-k1]

Biker's Nemesis said:

I've only asked how to turn wheelie control off. The nearest to having an accident on my MT 10 was caused by ABS at very slow speed where it activated causing my to run on over a junction.

So, for clarity were you ...

1. Approaching a junction at speed and left braking so late that you locked the front wheel up resulting in the the ABS saving you from a spill?
2. Travelling at a slower speed and snatched a fistful of front brake causing the front wheel to lock up resulting in the the ABS saving you from a spill?