Traction feedback device?

[CookMySock]

the more and more i think about this, to get ANY sort of accuracy you will have to have several models of the different suspension systems, and upon an install on ANY given bike, have to test and find the vaules to fit that model, and then retest as things like fork oil degrade etc etc....

No, you are over-thinking it. While it is true that the device would have to learn individual bikes, tyres, and road conditions ;

When you are riding, a little wiggle from the chassis might be related to a random patch of road surface, but as you observe earlier - now rider skill and the ability to not panic plays the major role.

The panic response from the rider, and the subsequent chassis flick and slide is abundantly clear to both yourself (even if you were following and watching said rider) and also clear to the digital signal processor monitoring the bike. Said nasty flick and slide will provide a very clear fingerprint to the DSP unit. Tyres, while pivotal to the safety of the whole operation, are included in the overall picture and not part of the individual analysis.

* Having said that ; Later on I cannot see why the device could not be adapted further to specifically target tyres and their behaviour, or for lots of other purposes too.

** The software on this thing will be a massive task,

Steve

[bogan]

firstly, i contest that fact through experience, and secondly, your original proposal was a meter to show how much grip was available to allow you to use 99% of it, which encourages users to do just that.

ok, yes oil can be a disaster, but many can handle it....

and get a rider with some offroad experience....


you've clearly already made up your mind on this, and the threefold system IS a vast improvement, so I think it's time we left you to it because clearly there isn't that much for KB to offer you until the next stage of the project.

good luck.

(oh, and all falsehoods pass through three stages too - ridicule, opposition, then through politics into law)

haha, touche.

Ok so some sbk buys can recover from oil slicks, many didnt as well. So for road riding i still say if you u see an oil slick, you gotta slow down. Ive said many times the rider is responsible for road surface hazards, i dont know how else to explain it, this system isnt designed to pick up on that!!

This may vastly reduce its useabilty, which is the point many of you are trying to make, but if we take corners with visible hazards out of the equation (by rider sees, rider slows) I think the predicated amount of traction, and the actual amount of traction may turn out to be fairly similar. But there is now way of knowing (beleive me when i say i wish there was) other than a lot of expensive testing, or write another modeling program (a lot of work, but costs nothing!).

[motorbyclist]

This may vastly reduce its useabilty, which is the point many of you are trying to make,

i was thinking about this on my commute this morning, and realised there IS a clear use for your product; rider training!

i am not talking about learning on the road, but as a training aid for when conditions ARE predictable which would mean riding schools and track days;
just a means of saying, "look, you've got far more room to improve that stop by x amount" or "yes, i can take that corner faster" or "yes, that is what we call a fuck-up; don't do it again"

so while i still say this product is a hazard on the road (and useless in a race), and if any accuracy is to be obtained it's cost will most likely be prohibitive. so get it accurate, reliable, and easily calibrated and sell it to riding schools where you know the bike, tyres and surface. They can afford it, they can oversee it's safe use, and it will help learners in their early stages to gain a bit of confidence

[CookMySock]

there IS a clear use for your product; rider training! [...] as a training aid for when conditions ARE predictable which would mean riding schools and track days; [...] just a means of saying, "look, you've got far more room to improve that stop by x amount" or "yes, i can take that corner faster" or "yes, that is what we call a fuck-up; don't do it again"

Yep.

and if any accuracy is to be obtained it's cost will most likely be prohibitive.

This is why it needs a GPL firmware - thats the quick way to get experienced mathematicians and coders all over the world behind it. It will be too massive a task otherwise.



Steve

[dpex]

Ive been toying with the idea of an electronic training device to help riders learn the limits by displaying the amount of traction being used in any given situation.

What it does is takes sensor data from the bike and works out all the net forces acting on the bike, it also calculates the weight distribution on each wheel. Using the cornering g-force, acceleration/decelleration forces, weight distributions, crest/dip forces, etc, it calculates the percentage of traction currently being used for each wheel and displays the highest one.

The value it uses for this maximum amount of traction is calculated using a number of constants which are programmed into the system, these are programed for each bike it is on. They include weight, rider and bike, tyre profile and friction co-efficient (can be found with two sets of scales).

What it cant do is sense the road conditions, potholes gravel, oil, wet, camber, the rider is responsible to back off if the road is slippery.

The display is currently under review, an LED bargraph is used in the current prototype, which shows the percentage scale in realtime, and records the max percentage used round a corner which it then displays after the corner for rider review. Other suggestions have been to use a beeper, beeps faster the more traction you use, and also to do datalogging in conjunction with GPS data.

The working idea of it is that it provides riders with a measure of the upper limit of traction they have available, (kind of like an intense track day would do, only without the practical experience obviously). For example, if a rider did a quick stop in 70m thinking he was using most of the traction, then put this gismo on the bike and found he could use a lot more, when he next did an emergency stop he could stop in 55m, thus being safer. There would be similar benifits while cornering and accelerating. The bit where this doesnt work is when the road conditions are poor, if he tried getting close to the gizmos idea of traction he would lock up, so the rider is still responsible for noticing porr road conditions and adjusting speeds and stopping force.

There is also the possiblity of a brak test to find the available traction, this would mean the rider rides along and locks up the rear wheel, the gizmo would then examine the deceleration needed and work out the force required to lock it up, and thus calculate the current road friction co-efficient, but again this would still not solve the problem of potholes, gravel spots, and oil slicks, though it would take care of different road quality and wet roads.

Your thoughts on this idea?

I think the concept has merit, but any form of visual read-out would be a no-no.

However, here's a thought. In gliders we have a dingus called a variometer. It tells the pilot when the aircraft is rising or falling. Although there is an analogue read-out of the rate of climb or fall, there is also an audio tone which is a flat-line buzz when the aircraft is neither rising nor falling. When the aircraft starts to rise, in any form of lift, the flat-line tone begins to beep (slowly if the rise-rate is slow). As the rate of rise increases so does the frequency of the beeps.

When the aircraft is falling (lack of lift) the tone goes from flat-line to a berurp. The berurp increases in frequency as the aircraft falls faster.

The benefit, as I'm sure you'll all appreciate, is that using such a tonal feedback for your device Head-phones in the helmet), one would not have to take precious micro-seconds of vision from the track ahead.

I've always been mildly impressed with my GPS which guides me here and there....'Turn left at the next lights. Take the 3rd exit from the roundabout. Perform a U-turn when possible, etc.

Personally, I think the makers could bump a bit of fun into such devices so that when you took the second, as opposed to the third exit, the voice might say, 'I said the 3rd exit, dick head'.

And so it could be with you device coupled to a tonal feedback system as described above.

You could have settings for "soft" cornering...you're in the burbs. "Medium and Death" for the roads. The latter two could provide the recipient who is going either too fast or too slow with an admonishment. "Too Fast!" the voice screams, 'You're about to die, Sucker!' Too slow and the voice intones, 'Hey, arsehole! Get a life.'

But when the rider gets the perfect balance of all parts of the turn the device might assert, 'WoooHooo. Have my baby. You is the king of the road.'

Stuff like that.

Yeah. I think the idea has merit, but visual feedback would be a no-no.

[Dare]

This is why it needs a GPL firmware - thats the quick way to get experienced mathematicians and coders all over the world behind it. It will be too massive a task otherwise.

Another way to get coders etc behind the system is to make it as adaptable as possible, this means using non-proprietry connectors, 12V in, modular system etc, etc. look at the arduino (www.arduino.cc) for a good example of this kind of system in action.

Imagine this, your initial system comprises of GPS, gyroscopic sensors, accellerometers and sensors for speed, throttle and braking input, using these sensors and some moderately smart coding we have a basic telemetry. this means we can see how smooth your inputs, leans etc were and even where you were when you did all that. This is very interesting stuff to look back on and see how fast you took X corner and how smooth you were when you did it. Bluetooth/USB would handle uploading this information to a phone/laptop with minimum of fuss. Onboard flash memory can handle firmware/software as people get onboard and find clever ways of interpreting all this data. Later modules could handle more and more aspects of the bike up to having a video display with PSI, Temp, G, and the list goes on.

As for having this information available at the time of riding, I believe no visual cue will be quick enough to react to. Think about it, from a visual input average human reaction time is roughly 0.2 seconds, that is more than enough time to lose traction past the point of no return. A noise is a good idea, with a sound scheme to alert the rider when different things are happening (record amounts of lean, for instance). A big part of learning to ride is to feel for what the bike is doing on a level that no dial or light is going to give you a sense of. The best they can do is give you a rough idea of when you should be on full alert to what the bike is trying to tell you.

Just my 0.02

[Lurch]

I think that rather than a dashboard display you would be better off developing a post-ride analysis for riders.

An onboard device that records the riding session in great detail including lean angle and forces which you download after the ride and gives you a corner by corner analysis viewable on a computer.

[bogan]

I think that rather than a dashboard display you would be better off developing a post-ride analysis for riders.

An onboard device that records the riding session in great detail including lean angle and forces which you download after the ride and gives you a corner by corner analysis viewable on a computer.

That would be great for track use but I was thinking the system would be more for road use. And after a long ride you would probably have forgotten what the corners felt like, it will still record it anyway but the dash indicator was to give the rider immediate feedback on the last corner taken. Ie go round corner fast, on the straight quick glance at indicator, says last corner was prolly a little bit sketch, dont go so fast the next time.

[Lurch]

Watch this guy get it wrong. And die. At 1:27 watch the brain-lock followed by the brains' refusal to bar push

The major problem displayed in this clip wasn't the SR involved it was the poor decision making that got him into that position on the road and at that speed.

(Sorry, bit off topic)

[t3mp0r4ry nzr]

Ive only read the first post, but immediate thought is info. overload and relevance of info.

on a bike your pretty busy anyway. so all this extra info., useful? if you cant take it in, what is the point?

if you can measure traction or lean, is that gonna make you a better rider?

[bogan]

Ive only read the first post, but immediate thought is info. overload and relevance of info.

on a bike your pretty busy anyway. so all this extra info., useful? if you cant take it in, what is the point?

if you can measure traction or lean, is that gonna make you a better rider?

taking in info is same as glancing down at speedo, so not information overload. I would say its pretty relevant, knowing how much traction you have left will make you less likely to exceed this level and crash

[aewilliam]

BIG PICTURE here...

If Bogan's little piece of wizardry is going to prevent, or contribute to preventing accidents, through TIMELY (i.e. PRE-EVENT...) information to the rider or the bike's ECU/braking systems) on your/my beloved two-wheeler, me thinks we ALL would say "Yes"?

I am not gonna even mention HOW it will work, but here i am talking about a WORKING DEVICE.

It sounds like it needs miniature time-travel technology to know what is gonna happen - either a couple of seconds for human reaction, or milliseconds for computer controlled reaction - in advance of your current riding situation/position.

But I dont care how it works now do I, just that it does.

The Government would say yes, contribute to R&D, provide production subsidies, and sooner or later, likely draft into law requirement for all new production bikes (or even cars?) sold in NZ, the installation of Bogan's device, just like ABS will likely be (or is it already?)

ACC and insurance premiums for bikers are given reason to drop (but whether they actually do is another question).

I will not only say yes, but also, Bogan, give me your bank a/c number and I will contribute to R&D if i can share in the financial gains.

But if all its gonna is say is "Dude, you f***ed up, and You are gonna be kissing tarmac 0.00005 seconds ago, and you have just paid $500 for me and I'm gonna get f***ed up as well, so you better hope your insurance is current"...

Then a "no" will be my answer.

[t3mp0r4ry nzr]

taking in info is same as glancing down at speedo, so not information overload. I would say its pretty relevant, knowing how much traction you have left will make you less likely to exceed this level and crash

bikes are unforgiving! if u overstep traction just for a fraction of a second, you may crash. you will not see this displayed on the monitor until you go to pick up the bike. relevance??

how about developing a plug-in ABS or traction control unit? prevent the crash from happening rather than letting you know that you overstepped traction AFTER the event.

just my thoughts as I am well versed in crashing!

good luck

BTW: maybe its just me, but from my experiences from racing means that I dont really look at the speedo, tach or even use mirrors when riding on the road (I look over me shoulder) and feel/hear the motor/judge my speed on other vehicles. less distractions on a bike, the better.

[Dargor]

Thats an amazing idea, but i would be very sceptical of the readings given from the method described.

[bogan]

bikes are unforgiving! if u overstep traction just for a fraction of a second, you may crash. you will not see this displayed on the monitor until you go to pick up the bike. relevance??

the monitor will display a 80% or similar just before you crash. And the bigger picture is you will go round a corner, the monitor says 80% of traction was used, so you wont try and go any faster round said corner.

how about developing a plug-in ABS or traction control unit? prevent the crash from happening rather than letting you know that you overstepped traction AFTER the event.

did consider that, but would have to manufacture sensor mounts individually for each fork and swingarm, and wheels etc, as well as all the liability issues etc

just my thoughts as I am well versed in crashing!

i appreciate the input, and hopefully some of you guys would be able to do some track testing when i eventually make a prototye.

good luck

BTW: maybe its just me, but from my experiences from racing means that I dont really look at the speedo, tach or even use mirrors when riding on the road (I look over me shoulder) and feel/hear the motor/judge my speed on other vehicles. less distractions on a bike, the better.

the user interface method can be whatever is required, visual, audible, reveiwable on comp after

BIG PICTURE here...

If Bogan's little piece of wizardry is going to prevent, or contribute to preventing accidents, through TIMELY (i.e. PRE-EVENT...) information to the rider or the bike's ECU/braking systems) on your/my beloved two-wheeler, me thinks we ALL would say "Yes"?

I am not gonna even mention HOW it will work, but here i am talking about a WORKING DEVICE.

It sounds like it needs miniature time-travel technology to know what is gonna happen - either a couple of seconds for human reaction, or milliseconds for computer controlled reaction - in advance of your current riding situation/position.

But I dont care how it works now do I, just that it does.

The Government would say yes, contribute to R&D, provide production subsidies, and sooner or later, likely draft into law requirement for all new production bikes (or even cars?) sold in NZ, the installation of Bogan's device, just like ABS will likely be (or is it already?)

ACC and insurance premiums for bikers are given reason to drop (but whether they actually do is another question).

I will not only say yes, but also, Bogan, give me your bank a/c number and I will contribute to R&D if i can share in the financial gains.

But if all its gonna is say is "Dude, you f***ed up, and You are gonna be kissing tarmac 0.00005 seconds ago, and you have just paid $500 for me and I'm gonna get f***ed up as well, so you better hope your insurance is current"...

Then a "no" will be my answer.

Basically its more of a training device than a pre-event system at this stage. As pre-event would require an accurate mapping of the road ahead, i have no doubt this can be done with computer vision but it would be far more complex than the system im currently proposing