Community Forums

Steven Groves · 04-13-2017, 10:48 AM

I mentioned autonomous driving to a sales guy at the local dealership and he suggested that "it'll never happen" - which I know to be bull-loney, but this is the guy who also suggested that a solar panel to keep batteries topped off was inane - even after I pointed out the cost of topping off a battery using fuel is $X and the cost of topping off using solar is $0 - but I digress.

It looks like the last post on this was in 2015, but I'm new to the site and still uncovering all kinds of smart people and threads that cover my interests; sorry if this is being covered elsewhere & open to being directed there.

My question is 'What's the state of the art now?', especially in light of the trite response from the dealer / sales guy.

I am on the look out not only for a manufacturer provided autonomous driving capability, but what kind of upgrades / kits are being developed to retrofit an existing motorhome.

Here's what I found recently -

I came across this post on Wired that suggested you could retrofit a semi for about $30K https://www.wired.com/2016/05/otto-retrofit-autonomous-self-driving-trucks/. I'm wondering why this kind of retrofit tech would not also be offered to RV owners?

It also looked like Perrone Robotics is beginning to offer a 'Drop-In Autonomous Kit' (DAK) http://www.perronerobotics.com/dak/ - which is not the kit I'd want, but it certainly confirms my suspicion that a retrofit is not outside the realm of comprehension.

Jalopnik cites a company (Cruise) as offering an upgrade for Audi in California, for only $10K! http://jalopnik.com/this-10-000-system-will-make-you-an-autonomous-car-be-1594843261

Anyone else following this topic? Love to hear what you're seeing!

Hufficon · 04-13-2017, 05:37 PM

Where did you find the free solar panels?

Sweetbriar · 04-13-2017, 06:12 PM

Most likely the first mass produced rig to see such will be a C class. Considering the integration required, (throttle, brakes, steering, and in particular the sensors) the C class cut away chassis coming out of Detroit or Stuttgart makes for a much more economical possibility. Plus in say 10-15 years the tech will have matured and costs will be affordable for the average buyer.

And for the OP, none of the links work.

Steven Groves · 04-13-2017, 07:01 PM

Quote:

Originally Posted by Hufficon

Where did you find the free solar panels?

OK, OK - agreed that the investment for the solar panel becomes a sunk cost. If I get a used rig that HAS them though... all other things being equal, I'd prefer solar panels... but again, we digress - seen anything cool about autonomous driving RVs?

Steven Groves · 04-13-2017, 07:09 PM

Quote:

Originally Posted by Sweetbriar

And for the OP, none of the links work.

Dang

- try these...

The content management platform of the IRV2 site is one I remember using from a decade ago - forgot how to use them; edited the wrong part of the URL.

Always willing to learn / re-learn! Just happy to be here...

THenman · 04-13-2017, 07:12 PM

Quote:

Originally Posted by Sweetbriar

Most likely the first mass produced rig to see such will be a C class. Considering the integration required, (throttle, brakes, steering, and in particular the sensors) the C class cut away chassis coming out of Detroit or Stuttgart makes for a much more economical possibility. Plus in say 10-15 years the tech will have matured and costs will be affordable for the average buyer.

And for the OP, none of the links work.

I would drop on down to B class for the first for your reasons.

Now this is just my $0.02 on the topic but no way I would trust a self driving vehicle. Even having had adaptive cruise on my car for a dozen years I am still alert to hit brakes or gas if a problem occurs (and they have). Also having had a mfg. business with around a hundred CNC machines I am here to tell you problems do occur where things do not move as expected. I just don't ever see where a vehicle will be prepared for everything. Will the time come I just sit there and watch waiting to do something if a problem occurs? Perhaps but I will never trust a computer and sensors 100%.

Tom

IC2 · 04-14-2017, 07:05 AM

Why would you ever want autonomous driving on an RV? You are supposed to using these rolling homes for enjoyment and that includes the driving experience. Secondly, and regardless of the fact that Elon Musk is pushing for his Tesla line to eventually become autonomous, how will this ever take on the task of unforeseen circumstances such as a child darting out in the road or more commonly, that semi or RV that just slowed down way more then expected for whatever reason. Or that roundabout where you are supposed to take turns but not all have gotten the message. I can have enough experiences while driving without a computer attempting to anticipate what that distracted person might and will do. As far as Jalopnik - they are cheerleaders for anything Tesla(Musk) or their parent, Cayman Islands registered Gawker, decides is 'news'.

Sorry about the rant but I have a problem with AI and its implications on life - mine in particular.

Steven Groves · 04-14-2017, 08:01 AM

The commentary is good. This is the future - welcome to it.

THenman - The improvements in autonomous drivers are light years ahead of where CNC is or ever was. I think they're pretty different, but the experience with how precise the CNC can be is good example of where it's headed. Trusting it 100%? I think I'm with you there, give me a few decades of success and then, well maybe 100% will be possible.

IC2 - I completely get that part of the appeal of an RV is the ability to have one of these behemoths under control of a mere human and yeah, the rant took away a little on the point you were trying to make. You could make an 'Autonomous Driving Rants' thread though and I'm pretty sure there would be a lot of participants.

I have a few comments I'd share in that discussion, but for now it'd take away from what I'm really trying to explore here, which is not 'if' this technology will become a reality, because it IS happening now and it'll be here within our lifetime (prob 10 years or so I'm thinking).

I'm looking more about 'when' & keeping an eye on the horizon for it.

Isaac-1 · 04-14-2017, 09:19 AM

I see this technology as too far off to predict how it will all shake out, anyone remember those low frame rate video phones that worked over normal phone lines AT&T tried to sell back around 1990. As I see it that is about the state of the industry today, sure it is getting better ,but people will expect facetime, and what is out there now is more like stop action animation. Sure lots of money is being dumped into this, so may happen quicker than the 15-20 years from AT&T's video phone to facetime, but guessing how it will all shake out is just as uncertain as forseeing the iPhone in 1990 before public access to the internet was even commonly available.

WhereBear · 04-14-2017, 03:03 PM

Hi All,

Driver Assistance and automated driving happens to be my day job, here is my 2 cents on where the technology is headed.

First off, there are many different definitions of 'self driving' and different levels of sensing are required depending on what it is you want to do.

The simplest form is highway single lane driving actively supervised by the driver. Basically a 'stay in this lane and don't hit the car in front' and if anything goes wrong control is instantly passed back to the driver. This requires a driver to actively monitor the system at all times and be prepared to take over at any time. This is basically the Tesla Autopilot 1.0 system and GM's upcoming Supercruise. This setup requires only a forward facing camera and radar unit.

The next level up is a system that has enough sensing and redundancy to allow the driver to not actively monitor the system, but be ready to take over given ~10 seconds warning. In order to continue operation for 10s after a single failure, you need redundant sensing, power, steering, brakes and control. AGAIN, highway only as this reduces the sensor set, though it's still pretty extensive.

To do full autonomy (no driver) and to do it on surface streets requires a huge increase in sensing and compute power as now the system has to deal with traffic lights, police hand signals, bikes, pedestrians, Emergency vehicles, school busses, etc. These vehicles will come sooner than most people think, but they will be owned by ride services companies and operate in a defined area that can be extensively mapped. Being part of a fleet allows constant maintenance and real time fleet learning. These will be confined initially to urban areas because that's where the economics make sense. If you have a car with $100k worth of sensors and computers, it needs to be running 24x7 making you money, not driving 30 miles to pick up the next fare.

Now, when will this stuff start showing up in motorhomes? Probably 5 years or so and I would agree it will be on Class B/C first because the OEM will have developed it and the MH coach manufacturer can just buy it. It will also be the first level of highway initially, but eventually the second level. Will be another 10-20 years I expect before you would be able to buy a fully autonomous (able to negotiate off highway areas) vehicle of any kind to have as your own. Motorhomes would be some time after that.

All I know is that I have about 30-35 years before my kids try to take away my license, hoping I can buy one before then!

Anyway, hope this helps the discussion.

Scott.

Ray,IN · 04-14-2017, 07:51 PM

Adding to post #7, if today's GPS mapping technology is not vastly improved first, this is something I'd never try.
For instance, about 7 miles from my house the state is rebuilding and completely redesigning an overpass bridge, it is so screwed up at present we who live here cannot traverse that bridge without a lot of forethought. It has a turn lane dead-ending into the left turn oncoming lane, but temporary lane markings indicate it continues onward to the opposite on-ramp.
It is not displayed on any GPS mapping program as under construction, let alone completely redesigned. I'd hate to be "riding" in a self-driving vehicle approaching that bridge. Until they solve that type problem it will never be safe.

Steven Groves · 04-15-2017, 05:48 AM

Quote:

Originally Posted by WhereBear

Hi All,

Driver Assistance and automated driving happens to be my day job, here is my 2 cents on where the technology is headed.

First off, there are many different definitions of 'self driving' and different levels of sensing are required depending on what it is you want to do.

The simplest form is highway single lane driving actively supervised by the driver. Basically a 'stay in this lane and don't hit the car in front' and if anything goes wrong control is instantly passed back to the driver. This requires a driver to actively monitor the system at all times and be prepared to take over at any time. This is basically the Tesla Autopilot 1.0 system and GM's upcoming Supercruise. This setup requires only a forward facing camera and radar unit.

The next level up is a system that has enough sensing and redundancy to allow the driver to not actively monitor the system, but be ready to take over given ~10 seconds warning. In order to continue operation for 10s after a single failure, you need redundant sensing, power, steering, brakes and control. AGAIN, highway only as this reduces the sensor set, though it's still pretty extensive.

To do full autonomy (no driver) and to do it on surface streets requires a huge increase in sensing and compute power as now the system has to deal with traffic lights, police hand signals, bikes, pedestrians, Emergency vehicles, school busses, etc. These vehicles will come sooner than most people think, but they will be owned by ride services companies and operate in a defined area that can be extensively mapped. Being part of a fleet allows constant maintenance and real time fleet learning. These will be confined initially to urban areas because that's where the economics make sense. If you have a car with $100k worth of sensors and computers, it needs to be running 24x7 making you money, not driving 30 miles to pick up the next fare.

Now, when will this stuff start showing up in motorhomes? Probably 5 years or so and I would agree it will be on Class B/C first because the OEM will have developed it and the MH coach manufacturer can just buy it. It will also be the first level of highway initially, but eventually the second level. Will be another 10-20 years I expect before you would be able to buy a fully autonomous (able to negotiate off highway areas) vehicle of any kind to have as your own. Motorhomes would be some time after that.

All I know is that I have about 30-35 years before my kids try to take away my license, hoping I can buy one before then!

Anyway, hope this helps the discussion.

Scott.

WhereBear, excellent... Just where I was trying to go... I KNEW it could not be that far off. Thanks amigo...

Oldone1 · 07-12-2017, 01:38 PM

think your right on track to let your mh put its self in a ditch

Steven Groves · 07-12-2017, 02:45 PM

Quote:

Originally Posted by Oldone1

think your right on track to let your mh put its self in a ditch

You're on track there I think, but only if I try to handle it myself.

I think the whole idea of this kind of stuff is to make the environment safer for not only the people in the unit, but also for those nearby.

I think the danger comes in the form of not adhering to manufacturer guidelines when they finally come out. The tragedies I've heard & seen in the news so far have ended up being attributed to operator error, not the errors of the technology behind it.

One other rumor I've heard, is that Tesla is looking at providing insurance on their cars - once they can be sure the car can drive better than the human.

Seems like a win for people in terms of safety & security!