Notice that there is a big push to so called "automatism" in "self driving vehicles".

Yeah, right!

This tech is totally dependent on an incredible number of in vehicle sensors and internet connected data.

How dependable are these sensors in less than optimal conditions (weather, corrosion of connectors, electrical or radio frequency interference, solar flares, etc.)

Yeah, not to mention the widespread failure of error free data over hardwired much less over the air communication channels!

The best of tech cannot deliver air bags that do not fail due to environmental conditions, engine control networks that do not require constant debugging, or even in vehicle entertainment or location devices that deliver consistent and accurate data.

Does your home computer consistently connect to the internet without error at optimal speed, does your GPS deliver millimeter accuracy, does your cell phone always deliver a reliable connection and superior voice quality no matter your location? Does the automated baggage system at your airport operate without delay or error? Does the light rail to that airport leave commuters stranded 30 feet above ground due to software errors, or require human intervention because the road crossing gates fail?

There has indeed been considerable advances in mechanical engineering, electronic controls and communication...

But I still carry a full set of tools in my vehicle, and I doubt that ANY damn sensor or computer will ever be able to react to the unforeseen better or quicker than the sound judgement of an experienced HUMAN operator!

 

Sorry, probably belongs in "Morning Coffee" thread.

 

Really the automated cars would probably be safer than all of these ******s driving and texting. Also, imagine how many "limp mode" or "cutoffs" there would be in a system like that.

 

For right now I am not seeing a fully automated system per se, more a hybrid one. I believe (I have yet to "drive" and autonomous vehicle) that the ones on the market now require some sort of driver input (turn the wheel slightly, etc) every few second or minutes (I forget which) just to make sure the person "driving" is still paying at least some attention and had not decided just to put their chair back and catch some z's or hopped over to the back seat to cook some stir fry or something.

This I do see value in, the vehicle can often "see" things we don't and can at least alert a driver and maybe take action where it deems an impending major catastrophe, and as such augments the ever growing lack of attention to check on bookface or zitter or whatever the cool kids are doing these days instead of actually driving that plagues our roads.

You can guarantee that they are collecting data en mass during this with the ultimate goal of full autonomy in the relatively near future. I am with sous, I definitely think it will ultimately and overall be a good thing, it will help congestion and will reduce accidents, unless someone modifies their system to drive like an a-hole of course lol.

I, and I believe most here will have some kind of push back to this new technology because we actually ENJOY driving our vehicles, the connection between man and machine will always win for me. But maybe a commuter self driver so I can get an extra 1 1/2 hours sleep on the way to work would be nice in 2025..... lol

I believe it is coming, but the lawyers, safety patrol, the government and of course the almighty insurance companies will make sure the system is flawless with multiple levels of failure protection/correction before they even think of marketing a fully automated vehicle to the public that would allow no oversight by a human being at all.

Just my 2 cent

 

I have no doubt the technology is close if not already here. Problem will be cost. Not only the cost to produce and install (vehicle and infrastructure) but the cost to maintain.


We already have vehicles capable of driving from the gate at an airport out to the runway, taking off and flying to another airport landing and driving to the gate all on automation. We don't use them to do all that but the capability is there.


The airports would have to install, maintain, and constantly test the sensors and equipment for ground guidance. It could be useful in places like London and San fran where dense fog is often encountered but the airports opt for a human driving a truck with lights for guidance. The cost and hassle overrides the usefulness of automation.


As for the airborne part our aircraft get programed daily by pilots with a flight plan that will operate from take off, climb, navigate across country with altitude and directional changes, to landing and rollout automatically. This costs the airlines tons of money but allows operation in weather and with increased time and fuel efficiency. The airline has to develop and maintain a FAA(government) approved autoflight program (office personnel, computer programs, etc), the aircraft have to be equipped, the pilots have to be trained, the maintenance techs have to be trained, the aircraft have to be tested frequently( monthly or sooner) and any sensor, servo, actuator, etc which shows any deviation replaced, pilots and maintenance techs require frequent recurrent training. It cost a lot but in business you spend money to make money so it's worth it.


There is also ground equipment at each airport that must be maintained and tested very frequently. Keep in mind airplanes fly on closed or regulated pathways that not just anybody can get on. Also while they are flying automatically there is always a human right there monitoring everything and ready to takeover if anything glitches. During certain phases of automatic flight the pilot actually has his hands on or hovering over the controls and is as focused as he would be if actually flying.


Now lets look at vehicles. They do not carry as many people and aren't flying but they are interacting with thousands of individuals and operating in a much less controlled environment. Even with a small car get 2500 lbs moving fast and then let it get out of control and you can have a disaster. Make that a truck or bus and more weight and more mass.


The cost part might be worth it for say trucking companies. I know I'm not comfortable with a trucking company trying to save money by getting rid of drivers if that means I have to be on the road with uncontrolled 18 wheelers, are you? Many trucking companies have trouble just passing current DOT safety inspections much less maintaining a highly sophisticated sensitive electronically controlled vehicle traveling right next to your car at 65 MPH with no one at the wheel. Imagine a fire in the electronics bay and the whole thing going haywire. No thanks.


What about passenger cars? We all know how much we hate going to the stealership right! So if the government were ever going to buy off on letting vehicles drive themselves on the open road how many regulations do you think will be involved. What if you had pay to take your car to the calibrator once a month and if any part didn't pass you had to have it replaced or your vehicles auto drive disabled.


Course we could build special automated only roads but then we are back to expense and hassle overriding benefit.


Will it ever happen, surely. In our lifetime, highly doubtful other than very small isolated places.


Just my .02 worth

 

Cant wait for this

 

Because advanced technology is so new to the industry I work in, my company had to invent a job title for what I do - "Technology Specialist". I have a ton of feedback on this, if the mods don't kick it for being off-topic - but I just got a call to go do my thing as I was typing this reply. I'll check back to see if this is still here.

 

So I'm an engineer, though I don't do anything with vehicles (other than tinker with my old Superduty)! When I was going through school back in '08-'09 all of this technology was already being considered. All very possible and likely within our lifetime (I'm optimistically assuming I'll live another 40 years or so)...
One huge advantage once the kinks are worked out is that "traffic" will be a thing of the past. They demonstrated even back then using simulation software that if cars were computer controlled that they would move extremely efficiently through big cities, even with ten times the volume we currently have.
Now that could mean cars are traveling 80-100 mph and missing each other by mere feet or inches... No slow drivers, no slow drivers in the left lane, no last minute lane changes, etc.
I'm sure there are drawbacks, cars would be required to be compatible and could be enforced much like emissions is now...so cost would be pushed onto the consumer, additional regulation, glitches in the software or catastrophic vehicle malfunctions to name a few, roadway infrastructure cost (if any) would be on the taxpayers..

 

Colorado Horseman started it off right. Ain't no automation in operation today that I would stake my life or my family's life on - at least not in the public streets. Automatic rail to the airport gate? Sure... nothing crosses its path. I can't count the number of times I've had to instruct users to reboot, restart the software, clean a camera or sensor lens, put a sensor or camera back where it belongs, clean the power supply, etc.... This is all stuff that causes phone calls before I even need to consider a visit.

Once it's determined something is really wrong, it can be recalibration, a broken wire, contaminated electrical contacts, a failed component, wrong supplies used, or some mechanic thinking he can make it better by modifying it. These are a few of the many examples of things that go wrong that I can do something about.

There are some things I can't fix without a phone call of my own... to the manufacturer. Only through trained observation can one pick up a bug in the software, at least to the point where you can say "When the left blinker is on and the car has to do a panic stop - the car stalls. It doesn't do it if the left blinker is not on." It's things like this that cause the need to spend time on the phone with software engineers and PLC programmers.

Then... on top of that... technology has outpaced the support infrastructure. We have mechanics today that service by trouble code. We "old timers" know from experience how flimsy this is. The technology industry needs honest-to-gawd troubleshooters that understand sensors, electrical, hydraulics, pneumatics, linear actuators/motors, imaging systems, software, AND general automotive mechanical. Walk down the street or into the grocery store on any given day. See a lot of those types wandering around?

 

We're all guinea pigs now. Should be fun to watch, from a distance.

 

It is hard to comprehend a self driving vehicle but there are some already out there being tested. I believe that Las Vegas has something being tested out there as we speak. My airplane mechanic's son works for Ford in Detroit and he reports that there is a fleet of them driving around the Ford plant out there. Not really doing anything other than driving around and doing as commanded.

 

Mechanic in an underground mine here. We have self driving haul trucks and loaders at work in the mine. The equipment runs in between shifts while blasting is taking place and no one is underground. The loader drives itself to the pile of rock and an then an operator at a control station outside operates it long enough to fill the bucket with rock then pushes a button and it drives itself to a dump point, dumps the load and drives back to the pile. The truck will back itself under a chute to get loaded with rock then proceed outside where it waits for an operator to drive it to the dump station then send it back for another load.

 

Well I'll be damned, that is pretty cool if you ask me.

 

That's pretty straightforward stuff... A Roomba can navigate your house, and clean the floor as it goes. Technology has gone waaay past that for many years. You've heard of facial recognition software... but how about scar recognition software? If a person has a scar, one might look at it and recognize if it's from a heat burn, a chemical burn, a birthmark, an illness, a skin condition - and which skin condition.

Now... take an apple (potato, cherry, blueberry, whatever) that has a mark on it. There are systems that look at the mark and determine whether it's an acceptable mark or not. I'm not talking about just the size of the mark, I'm talking about the characteristics. In the produce world, a bank of computers and cameras look at each mark/blemish on the product, identifies the characteristics of the blemish (limb rub, worm, bitterpit, thrip, bruise, soft spot, etc...) and sorts that product to it's appropriate grading. Oh... and it hands off one of these decisions every 30 milliseconds - that's over 33 per second... per lane (on a cherry machine). One of our sites has two 16-lane machines, and we have other machines like this at different sites.

The other amazing technology is the one that determines the internal quality - firmness, internal browning, sugar content, watercore, etc... without touching the fruit. That one hands off over 10 per second per lane. If it were cherries, that would be too slow - but that one is apples... so we have 20 lanes of those.

 

It is pretty cool to see a 150,000LB machine operating on its own in the confined space of an underground mine.