Discussion
PanzerCommander said:
There is no reason why a system like this on a car (not a joystick controller but full drive by wire) should be unsafe, the firmware would have to be tested to a very high level because it is a safety critical component.
I agree, it will most likely have an extremely high safety rating, however, it will never be as safe and failure-proof as a mechanical steering rack in my opinion. All it takes is for it to fail under unfortunate circumstances before people are (quite rightly) a bit miffed that a member of their family / friend was in a fatal accident because a component of their fancy new fully electric steering fails.PanzerCommander said:
The only problem I can think of with a system like this is the loss of steering feel, which is a major part of driving, it lets you know all sorts of details about the road surface. But you can be that they will counter this with, it will have artificial feedback!
Even if this was the only problem (which I'm not convinced it is), then why would you actually apply the electronic steering? What benefits would it provide? 
BriC175 said:
Even if this was the only problem (which I'm not convinced it is), then why would you actually apply the electronic steering? What benefits would it provide?
I suspect the big one in the long run is packaging - no need to find a path through the engine bay for the steering column. RenesisEvo said:
I would hope (because I don't know for certain) that bump steer wouldn't be introduced into the suspension geometry in the first place; as far as I understand it's a direct function of the geometry of the suspension and its motion. It might be acceptable to eliminate bump-steer or roll-steer via a steer-by-wire rack, but to me it seems a very backwards way of going about designing the suspension. Rather than cover up a flaw, why not eliminate the flaw in the first place? Unless there is some significant advantage to be gained by doing so.
Most, if not all road cars are designed with bump steer built in for various reasons. It's only really undesirable in race cars, but even then it can still be useful.sjg said:
The Nissan system as linked has three ECUs - it would take all three to fail for the clutch in the steering column to need to be engaged.
If it's like the ones in aircraft, the ECUs monitor each other, and if one doesn't agree with the other two, it flags a warning and gets ignored (and possibly restarted).
So lets say you have 3 ecus. Are they from the same batch? Would they be running the same software? Do you have 3 sets of sensors or just one? What about the software testing? how rigorous would that be? What about faiilure? How much for a new one? what about the drive motors as I assume it would have to have at least 2 of those to guarantee safety and tbh they're more likely to burn out than the ecus? how much for one of them?If it's like the ones in aircraft, the ECUs monitor each other, and if one doesn't agree with the other two, it flags a warning and gets ignored (and possibly restarted).
Massive introduced complexity just so some designer can put the driver on the back seat? So much for Occam's Razor.
kambites said:
I suspect the big one in the long run is packaging - no need to find a path through the engine bay for the steering column.
Not to mention, this is of course from a Japanese company, Right hand drive home market.
No longer a headache in design between LHD and RHD cars, under the bonnet would be exactly the same for both. But that would only become evident one they had got the stage of not having the clutched fail safe column.
kambites said:
I suspect the big one in the long run is packaging - no need to find a path through the engine bay for the steering column.
I'm not really sure that packaging is an issue though? Engines and steering assemblies are fairly generic shapes, and the routing of steering columns and racks seems to be fairly standard and fit together well. The extra space is only a benefit if you plan to use it for something.renrut said:
So lets say you have 3 ecus. Are they from the same batch? Would they be running the same software? Do you have 3 sets of sensors or just one? What about the software testing? how rigorous would that be? What about faiilure? How much for a new one? what about the drive motors as I assume it would have to have at least 2 of those to guarantee safety and tbh they're more likely to burn out than the ecus? how much for one of them?
Massive introduced complexity just so some designer can put the driver on the back seat? So much for Occam's Razor.
Exactly! The amount of extra engineering and cost just doesn't seem to make sense!Massive introduced complexity just so some designer can put the driver on the back seat? So much for Occam's Razor.
roverspeed said:
Not to mention, this is of course from a Japanese company, Right hand drive home market.
No longer a headache in design between LHD and RHD cars, under the bonnet would be exactly the same for both. But that would only become evident one they had got the stage of not having the clutched fail safe column.
Do you not think the design and implementation of a whole new steering set up which replaces a system that's been reliably used since the begining of combustion engined automobiles would be more of a headache?
BriC175 said:
A headache? Most cars shells are a mirror image down the middle at the bulkhead as far as I'm aware, so that modifying them for LHD / RHD is just a case of running two different batches of steering related components and other differences (dashboard, seats, etc which will all need to be designed and manufactured regardless of steering set up). Something that's been done for many years, and I'm sure car manufacturers are fairly proficient at doing.
The bodyshell might be symmetrical but I've yet to encounter an engine bay on any vehicle that is. Exhausts, intakes, gearboxes, brake boosters and systems, ECU's, batteries, you would be amazed how much has to be different between most LHD and RHD vehicles.
Cars that have relatively low volumes of RHD (e.g. French) tend to have some fairly horrendous systems like metal tubes running behind the dash attached to the brake pedal.
BriC175 said:
I'm not really sure that packaging is an issue though? Engines and steering assemblies are fairly generic shapes, and the routing of steering columns and racks seems to be fairly standard and fit together well. The extra space is only a benefit if you plan to use it for something.
I guess the extra space would be removed to make the car smaller lighter and more compact, or moving other components into that space (not sure what) to improve interior or boot space for instance. It is also less crash testing as the collapsible steering column that would be tested now no longer exists – money saved. With EV’s it is more space for batteries or less weight which improves performance.It would also centralise the steering wheel for the driver, In another thread a complaint was levelled at the Mustang model that I drive of the steering wheel being off centre - it is a bit (not that It affects the way you drive to be honest as it isn't a ridiculous amount) but then the column has to worm its way past a wide engine and a set of exhaust headers, with a totally electric system the wheel would be perfectly central. So for cars with big engines it solves that issue.
As others have said it also removes the LHD RHD parts issues, the only difference would be the dashboard and brake and clutch servo position (assuming you weren’t using electronically actuated brakes and clutch) the rack would be the same in all cars, thus saving money on parts, because EPAS is already available it would shave a few pounds of the cars front end weight as well.
But the whole lack of feel touted in that article is to me a bit dodgy and borderline dangerous because there are enough “white goods” driving lemmings on the road that this system would just give another level of invulnerability to.
Not that I'd want a system like that in my Mustang because I like the real feedback from the wheels transmitted down the steering column
I dont think that generally in most situations on the road that the car is actually the problem, we can argue about the semantic details of the system but I still reckon Nissan can make this more reliable than the driver, we accept all manner of risk on the roads every day, this is one of the smaller ones.
I suspect when driverless tech reaches viability it will get legislated against before it gets to market or possibly there will be an accident involving fatalities, the Daily Mail will go into overdrive "Oh the Humanity", pictures of victims, bent metal and general wailing and gnashing of teeth, we do however accept that every year 2000 odd people will die on our roads, if driverless tech meant this was reduced to 1000 would we still want it banning because it causes deaths or is it just the fact us squashy, fragile and sometimes stupid creature insist on hurtling round in metal boxes at speeds we arent designed for ?
I suspect when driverless tech reaches viability it will get legislated against before it gets to market or possibly there will be an accident involving fatalities, the Daily Mail will go into overdrive "Oh the Humanity", pictures of victims, bent metal and general wailing and gnashing of teeth, we do however accept that every year 2000 odd people will die on our roads, if driverless tech meant this was reduced to 1000 would we still want it banning because it causes deaths or is it just the fact us squashy, fragile and sometimes stupid creature insist on hurtling round in metal boxes at speeds we arent designed for ?
In fairness I can see the point of such a system, and realistically it has the potential to work well, perhaps even to be better than existing electric setups which could be considered a bit of a compromised affair.
However I suspect that, to justify it's existence, engineers will try to be too clever with it and it'll end up being awful for at least the first few incarnations.
However I suspect that, to justify it's existence, engineers will try to be too clever with it and it'll end up being awful for at least the first few incarnations.
Driverless tech won't be a sudden change, just a series of small steps.
We're already at the point where adaptive cruise control is being fitted to mid-range, mainstream cars. We have systems that can read road signs (currently used to remind drivers of the current speed limit). We have collision mitigation braking systems, if the car thinks it will hit something, it will intervene if necessary. We have lane departure warning systems, and in some cases, active lane keeping systems (eg. on the new E-Class).
Put all that together and you have systems capable of handling most motorway driving, and probably safer than most drivers can. What's more of a challenge are the situations that involve more interaction with other traffic, and of course the legislation around what constitutes control of the vehicle.
We're already at the point where adaptive cruise control is being fitted to mid-range, mainstream cars. We have systems that can read road signs (currently used to remind drivers of the current speed limit). We have collision mitigation braking systems, if the car thinks it will hit something, it will intervene if necessary. We have lane departure warning systems, and in some cases, active lane keeping systems (eg. on the new E-Class).
Put all that together and you have systems capable of handling most motorway driving, and probably safer than most drivers can. What's more of a challenge are the situations that involve more interaction with other traffic, and of course the legislation around what constitutes control of the vehicle.
The Wookie said:
Most, if not all road cars are designed with bump steer built in for various reasons. It's only really undesirable in race cars, but even then it can still be useful.
If you don't mind, can you elaborate on these reasons? I've not encountered any reasons how or why it might be useful, and as my background in suspension design is race-car orientated, and as you say it's generally undesirable (I may have made a slight miscalculation with the bump-steer on a Formula Student car - had to move the upright track-rod mounting some distance to fix it).ohtari said:
Have none of you lot ever driven a tractor? Hydraulic steering, no feedback, no bumpstops on the steering wheel. Now that is interesting to thread roads with! Admittedly, being able to twirl the wheels from lock to lock with a flick of the wrist is useful in a field, but at speed, among traffic, not fun.
Tractors still have some feedback and bumpstops, and are getting a bit better at having some feel, however, a shift in a tractor can be as long as 20 hours, which you dont really want much feedback on while driving over ploughed land. Tractors designed for the road (I.e. fastracs and MB tracs) have much more feel. tractors also have very large front wheels which would make it difficult to turn without a lot of assistance - and indeed you can still steer with no power in a hydraulic system, its just very slow. - hence why tractors are limited to 20MPH in most cases, as you wouldnt want to be traveling at 60 and suddenly not be able to steer. I wonder if this new system will have a speed limit?personally i prefer no PAS at all and no stupidly over served brakes in my cars, but then im an idiot.
roverspeed said:
kambites said:
I suspect the big one in the long run is packaging - no need to find a path through the engine bay for the steering column.
Not to mention, this is of course from a Japanese company, Right hand drive home market.
No longer a headache in design between LHD and RHD cars, under the bonnet would be exactly the same for both. But that would only become evident one they had got the stage of not having the clutched fail safe column.
Triumph Man said:
So, and apologies if it has already been said, if there is still a steering column, what the hell is the point of an SBW system? Just another useless piece of technology for technology's sake. Bah humbug.
There is in this case because they think it's needed to make drivers feel safe. They have stated that the column is unnecessary and they will be aiming to remove it in future cars, though.kambites said:
Triumph Man said:
So, and apologies if it has already been said, if there is still a steering column, what the hell is the point of an SBW system? Just another useless piece of technology for technology's sake. Bah humbug.
There is in this case because they think it's needed to make drivers feel safe. They have stated that the column is unnecessary and they will be aiming to remove it in future cars, though.
RenesisEvo said:
If you don't mind, can you elaborate on these reasons? I've not encountered any reasons how or why it might be useful, and as my background in suspension design is race-car orientated, and as you say it's generally undesirable (I may have made a slight miscalculation with the bump-steer on a Formula Student car - had to move the upright track-rod mounting some distance to fix it).
Well to start off with, instead of toe-out or toe-in with bump, we typically term it understeer or oversteer, which should give you an indication of where I'm going. Toe in in bump is termed as oversteer on the front axle or understeer on the rear axle, and toe out is vice versa.You can use it to create a passive steer effect with roll and alter the balance as the car loads up, so for example if you set a car up with understeer on the front axle, it will tend to steer the wheels to the outside as the car rolls which is good for stability.
Oversteer on the rear axle also generally tends to improve yaw damping as the rear steer helps the transition into oversteer while removing the rear steer if the axle loses grip and sits up. Instead if you have understeer on the back you might increase the amount of natural understeer the vehicle has, but if the rear suddenly loses grip then the car will sit up again, and all of that rear steer will disappear, suddenly pinging the rear wheels more towards the outside of the corner and making the car yaw more aggressively!
Generally it also tends to be considered along with things like ackermann and compliance, which you don't really have with a ball jointed race car.
No doubt you've found it a pain with toe change under braking and acceleration, but it can still be useful. For example using oversteer on the rear axle of a front wheel drive car can help corner entry stability, mid corner understeer and corner exit as it's giving toe in while diving on the brakes, steering the rear of the car towards the outside of the corner in roll, and giving a bit of toe out when it's squatting under acceleration.
That's more or less the thrust of it, there's probably more to it, but unfortunately in the year I've been doing this job I haven't had much of a chance to play with it on road cars yet as the need hasn't come up and I'm learning as I go.
But next time you're turning a road car into a race car, don't think the designers were incompetent while you're shifting the steering rack because it's got a load of bump steer built in, there's usually a good reason for it!!
Edited by The Wookie on Thursday 18th October 16:16
Gassing Station | General Gassing | Top of Page | What's New | My Stuff