The Market Ticker
Rss Icon RSS available
You are not signed on; if you are a visitor please register for a free account!
Comments on Bye-Bye Truckers....
User: Not logged on
Top Forum Top Login FAQ Register Clear Cookie
Showing Page 2 of 2  First12Last
User Info Bye-Bye Truckers.... in forum [Market-Ticker]
Flappingeagle
Posts: 2969
Incept: 2011-04-14

Report This As A Bad Post Add To Your Ignored User List
For awhile will there be any market for a driver/driver-less hybrid? The driver oversees loading, seals the trailer, and then drives until out on the interstate. Then the computer takes over and the driver rests/sleeps until the truck gets it to the destination. At that point the driver negotiates it thru the town to the unloading dock.

I can see that happening a couple of ways and not happening a lot of ways.

Flap

----------
Here are my predictions for everyone to see:
S&P 500 at 320, DOW at 2200, Gold $300/oz, and Corn $2/bu.
No sign that housing, equities, or farmland are in a bubble- Yellen 11/14/13
Trying to leave the Rat Race to the rats...
Emg
Posts: 297
Incept: 2012-11-20

Canada
Report This As A Bad Post Add To Your Ignored User List
"The self driving cloud system goes offline several days in a row and their socialist utopia cities start looking like mad max"

I suspect one reason the left want self-driving trucks is so right-wing truckers won't be able to starve the cities by refusing to drive.

Either way, it's inevitable in the long term; even if, in the short term, they have to keep drivers on board to take over in situations the truck can't handle. That would still potentially allow the trucks to run 24 hours a day since the driver won't have to be working most of the time.
Tickerguy
Posts: 156031
Incept: 2007-06-26
A True American Patriot!
Report This As A Bad Post Add To Your Ignored User List
@Flappingeagle -- Due to the existing installed fleet of cabs (LOTS of them) what you're almost-certainly going to see will by a hybrid model.

Driver takes truck from terminal to a lot right near the on-ramp of the interstate and drops the trailer, where there is a waiting AV cab.

Said AV cab connects (perhaps with the driver plugging in the air hoses, since otherwise you have to figure out some sort of autonomous-friendly retrofit to the existing connections for those) and off it goes.

At the other end it pulls into a lot right off the interstate and waits for a driver-equipped cab to come and "release" it (e.g. disconnect said airhoses, etc.) at which point it either picks up another trailer going somewhere else, waits for work or departs as a dead-head to wherever it is next needed. The driver takes the trailer the last few miles to the destination.

----------
Winding it down.
Mangymutt
Posts: 865
Incept: 2015-05-03

Vancouver WA
Report This As A Bad Post Add To Your Ignored User List
Whitehat -
Quote:
the roadway must have a standardized way of interacting with the vehicle.


Karl -
Quote:
Driver takes truck from terminal to a lot right near the on-ramp of the interstate


and

Quote:
At the other end it pulls into a lot right off the interstate


Just think how much this will save if the cost is pushed onto the tax payer.

The money Amazons saves by having the tax payers pay for this can then be funneled to those who are out of a job and universal payed income can be implemented. This means MORE people will have money to pay even MORE taxes. Which means MORE tax payer money.

Damn AOC is a genius.

I now also believe the Greed.....errrr sorry Green Deal will do wonders for our country.



Elkad
Posts: 524
Incept: 2009-09-04

Report This As A Bad Post Add To Your Ignored User List
Another intermediate step I've seen mentioned is the road train. I think Volvo was working on it.
One human driver in the lead truck. 1-5 computer-driven trucks following him to the same destination.

You could further hybrid that with a system where the human does the dropoff, pickup, fueling, and town navigating, and when he gets to the interstate he links up with another company truck, puts his truck in auto-follow mode and goes off-shift. Now the auto-trucks don't even have to handle the drop&hook themselves. Since you have retained a driver, you can have an alarm in the sleeper where the truck can call for help due to bad weather, detour, etc.
You'd refit your newest power units for combo mode, while replacing your oldest ones with fully automatic units (maybe with a bare minimum seat for emergencies).
Emg
Posts: 297
Incept: 2012-11-20

Canada
Report This As A Bad Post Add To Your Ignored User List
"One human driver in the lead truck. 1-5 computer-driven trucks following him to the same destination."

How are cars supposed to join or leave the highway with six trucks driving nose-to-tail in the slow lane?
Asimov
Posts: 110549
Incept: 2007-08-26

East Tennessee Eastern Time
Report This As A Bad Post Add To Your Ignored User List
Emg: Just swerve at them and act like you're going to hit them - they'll move.

If they don't, just sue the everloving **** out of everybody related to their functioning. If you survive.

----------
It's justifiably immoral to deal morally with an immoral entity.

Festina lente.
Flaps10
Posts: 6980
Incept: 2008-10-17

PNW
Report This As A Bad Post Add To Your Ignored User List
EMG,
Well, comrade, you won't be merging into traffic.
Reluctantdebtor
Posts: 307
Incept: 2010-03-05

ohio
Report This As A Bad Post Add To Your Ignored User List
The temptation for displaced, resentful ex-truck drivers to cause mischief for autonomous trucks will be great. There are too many obvious ways to do that, and idle hands are the devil's best girl. Bonus: No attempted murder rap, which is always a risk when sabotaging the routes of conventional trucks.
Supertruckertom
Posts: 2535
Incept: 2010-11-07

USA
Report This As A Bad Post Add To Your Ignored User List
Now you know why my goal is get get the house paid off pronto.
Can still make $27 per hour driving a forklift while I get retrained to do CNC.
Somebody has to make the guns for the Resistance.

Company just raised mileage py to 69 cents per mile.
My route is 2750 miles per week.
3 weeks paid vacation.
Health Insurance is only $65 a week for a family of 5. Above that it goes to $100 per week.
Cigna Network $5k deductible. Flex Savings also.
50% 401k match
Non-Union LTL company.
Teamsters are making more, finally.
City P&D route drivers will still be around.
Line Haul terminal to terminal operations at night will be automated first especially on longer routes with filled to capacity trailers.
We already run double 48 footers on the Florida turnpike.
The Truck staging areas are already built to couple them together at each major entrance to the Turnpike.
UPS and Fed Ex will be the first adopters. Old Dominion probably next.
Where I work the owners are a bit more conservative in rushing to adopt new technology.
Privately held and debt free so they don't like putting money into unproven technology.
They actually rebuild hundreds of drive trains a year and install them into new glider frame/body kits to avoid the DEF.

I'm taking suggestions for my second career.


----------
Preparing to go Hunting.
Jazen
Posts: 3828
Incept: 2007-07-17

****cago
Report This As A Bad Post Add To Your Ignored User List
If you have CNC experience, I would look into CNC programming if I were you. I like to learn it for fun/hobby, but it could be quite.useful to know

----------
I hate our Government, but I still love America.
Tonythetiger
Posts: 5
Incept: 2019-01-27

Fort Walton
Report This As A Bad Post Add To Your Ignored User List
LIDAR will certainly outperform passive optics, but still won't deal with bad weather sufficiently well to make the grade. Some google based facts I dug up:

Stopping distance of Tractor-trailer at 65 mph on dry pavement ~= 160 meters
Stopping distance of Tractor-trailer at 55 mph on dry pavement ~= 102 meters
Stopping distance of Tractor-trailer at 40 mph on dry pavement ~= 51 meters

https://www.udot.utah.gov/trucksmart/mot....

Effective range of LIDAR:

Clear weather: = ~ 150 m
Dark and clear: = ~ 150 m
Heavy rain, snow, fog: = ~ 50 m

https://cleantechnica.com/2016/07/29/tes....

Mix in wet or snowy pavement and stopping distances get much larger.

So you'd probably be looking at speeds between 30 mph and 40 mph, best case, in inclement weather.

Based on studies done by Oakridge National Labs, tractor-trailer fuel economy (FE) on flat terrain is maximized around 60 mph. FE is around 30+% lower at 50 mph. They didn't measure at 40 mph, but it's probably worse than 50.

https://info.ornl.gov/sites/publications....

So expect the fuel costs to increase 50% or more when relying on LIDAR in bad weather.

Radar doesn't have as significant a reduction in bad weather, but it has lower resolution. At Ku band (17 GHz roughly) a 3 inch diameter antenna has a 4.5 degree beamwidth. At 150 meters (max range) that produces a detection zone that is 38 feet wide. Any non-moving object, located anywhere in that 38 foot wide swath, looks the same to the radar.

A stationary object 18 feet to the side of the rig could trigger it to slam on the brakes, even though it might easily drive straight ahead without impact.

Of course you could increase the size of the antenna to make the beam smaller, but then the cost of the radar goes up.

OR, you could try going to Ka band at 35 GHz, but the absorption in fog, smoke, and rain increases significantly and max range decreases.

So Radar is no slam dunk either. The sensor issue is quite challenging without some sort of man-in-the-loop backup system. Design for nominal conditions might be reliable, but the environmental extremes are a different story, IMHO.
Elkad
Posts: 524
Incept: 2009-09-04

Report This As A Bad Post Add To Your Ignored User List
Supertrucker's post above and some math.
2750*.69*50 weeks= $80k or so. Plus their portion of insurance, 401k matching, employer taxes, etc. So a minimum of $100k/yr to keep a truck rolling maybe 40% of the time (truckers work a lot of OT).

If refitting for self-drive cost $200k and got the truck to 80% usage (still needs loading time), they've got a $200k/yr team of employees paid for in the FIRST year. With a 10yr lifespan on even the drivetrain.

New tractors do even better. They can omit most of the creature comforts and run tiny day cabs with a jumpseat for mechanics/emergencies.

I wouldn't count on forklift jobs being around much longer either. It has it's own machine learning problems, but not any harder than driving. Just a smaller market. Basically the same software and sensors will cascade down to forklifts (and lawn mowers and everything else) eventually, as economy of scale from installation in millions of cars brings the prices down on sensors and processing. When you can slap a sensor pack on a forklift for $20k, the job of the full-time forklift driver, even at minimum wage, is gone too.
Gonewest
Posts: 70
Incept: 2015-02-26

PacificNW
Report This As A Bad Post Add To Your Ignored User List
One word: train ???
Elkad
Posts: 524
Incept: 2009-09-04

Report This As A Bad Post Add To Your Ignored User List
Tony, those braking distances, like every government publication on the subject, are pretty pessimistic (by about 50%).

I don't think there is a car made that can't do 60-0 in under 140'. Even a softly sprung pickup with offroad tires (Ford Raptor) is 150'. Sports cars on summer tires dip under 100'.
The semi numbers have similar problems.

A computer isn't going to have a reaction time (well, one best measured in milliseconds), and won't waffle about "hard stop" vs "panic stop" like a human does either, so it shaves some time off real-world as well. And it damn sure won't ever be driving distracted.

So I think Lidar will manage just fine, especially when combined with front-facing cameras and radar.
Tickerguy
Posts: 156031
Incept: 2007-06-26
A True American Patriot!
Report This As A Bad Post Add To Your Ignored User List
The amusing part of the LIDAR discussion is that as I noted it can see as well as you can, roughly.

Machine vision in the visible spectrum is less-acute then humans. Period. Even very high-dollar camera sensors (I own a couple) can't compete in terms of resolving power with the human eye -- and it's not a close call either. Under bad conditions this matters a LOT.

But -- add IR and millimeter-wave RF (radar) sensors and the game changes. You don't need to know WHAT something is at longer ranges (e.g. can you run it over) only that it obstructs the path, as that informs maximum speed you can reasonably sustain under conditions.

Note that braking distances include reaction time (at 60mph you cover 88'/sec) and human reaction time to visual stimulation is at best ~250ms ASSUMING NO DISTRACTIONS AND A ~20 YEAR OLD BRAIN. In the real world operating a car measured reaction times, including the decision to act, range from ~750-3,000ms, with most accident reconstruction teams using around 1,500ms for a "reasonable" reaction time to be expected. In other words, if, using 1,500ms reaction time you would have struck whatever you're likely to escape at least criminal and likely civil liability.

Well, a computer essentially eliminates that; the reaction time expected from these systems is measured in tens of milliseconds. At 60mph the human gets over 100' of travel before being expected to react to the visual stimulus (a third of a football field!) while the computer will react within a few feet -- and win every time.

The issue with the computer will be mis-classification and when that happens the maker will get tagged with liability. The open question is going to be where the legal boundaries are in the decision matrix and who's going to set them (e.g. what defines "deciding to run that over" is a LEGITIMATE action.) It's not as simple as it sounds -- if there's a human in the road can the computer hit him on purpose if the alternative is to smash into the vehicle(s) to the left or right?

----------
Winding it down.

Nadavegan
Posts: 146
Incept: 2017-05-03

The South
Report This As A Bad Post Add To Your Ignored User List
There will be a cottage industry of pilots around each offloading and fueling stop. The lots where the auto rigs pull over will have to be managed. Some states will require a living presence in the cabs. Etc.

Truck stops aren't going away. Long haul drivers may, but I bet net-net this ends up being a cost-shifting exercise in some large fashion, as personnel requirements will just be redeployed.
Elkad
Posts: 524
Incept: 2009-09-04

Report This As A Bad Post Add To Your Ignored User List
Nadavegan,

Except hiring a guy to jockey trailers around a lot and sleep in his own bed every night is not a $80k/yr job. Doesn't matter if it's a yard or a fuel station. If he handles 4 trucks an hour, and the average truck run is 10 hours long before a human needs to interact with it (wild-assed guess), he's replacing 40 drivers.

The non-union coal mines around here run guys who haul from the minehead to the rail spur. No CDL required because it's all private property, and it pays $30k/yr for bouncing up and down steep gravel "roads" all day and getting black lung. Moving trucks in and out of the oil change pit or around a yard won't pay any better.

You've just reduced labor costs by 99%. The states that require a qualified driver onboard will fall in line within a few years, and see a huge influx of qualified drivers in the meantime, which lets them cut pay drastically.

Hell, they can probably hire someone overseas to videogame some of the "human required" stuff for a couple bucks a day. We remote control drones in Iraq from the US, this is no different.
Gauntlet33
Posts: 87
Incept: 2009-03-30

Report This As A Bad Post Add To Your Ignored User List
Elkad: The states that require a qualified driver onboard will fall in line within a few years, and see a huge influx of qualified drivers in the meantime, which lets them cut pay drastically.

Even if the states don't agree to VOLUNTARILY fall in line, they may be required to do so as a violation of the Commerce Clause of the US Constitution as placing an "undue burden" on interstate commerce.
Hstella
Posts: 754
Incept: 2009-08-18

Colorado
Report This As A Bad Post Add To Your Ignored User List
Even if the roads are a disaster (I-70 is usually FUBAR in any snowstorm), the self-driving truck does have the option of just pulling over and parking until the conditions are better. But if the rig is otherwise capable of running 24/7, even a significant time loss from snow can be more than made up in comparison to a human driver needing 8+ hours of downtime/day. Crappy weather happens all the time, but truly awful weather (the kind that jackknifes human-driven rigs, too) certainly is not a 1/3 event. In CO we have a bad driving day the day of a 6+ inch snowstorm, a slushy but manageable mess the next day, and clear/dry roads the third.
Wa9jml
Posts: 319
Incept: 2017-04-29

DeKalb, Illinois
Report This As A Bad Post Add To Your Ignored User List
I have a decent knowledge of millimeter wave radars, thanks to being a project engineer on a traffic detection radar for the US DOT, some years ago. I built up a K-Band, 24 GHz Doppler radar with a 50 milliwatt Gunn effect oscillator, and a mixer/detector unit. I had it mounted behind a 2 foot dish with a bent WR-42 section of waveguide with a choke flange at the focal point of the dish. That antenna was designed by one of the best antenna engineers in the field. I took it up on the roof of our plant, and had radio contact with a few of my fellow engineers. I had learned in my military ground surveillance radar experience that radars see things very differently from our optical eyes. There was a van parked across the street, so I pointed the radar at it, and then directed an engineer to stand along the side of that van. Human beings are absorbers of 24 GHz signals, so as he walked along side that van, I realized that the only part of that van that was radiating back to me was the running board. There was a jeep parked down a side street, and the only thing that reflected back the signal to the radar was the license plate on the spare tire mounted on the back. Some time later, I did some of the research work with the late George Gill, who was a noted millimeter wave antenna expert. I helped him out on his design of a compact antenna range that was implemented at 94 GHz, and then scaled down in frequency. I learned that a big problem with compact antenna ranges was radiation of the signal from the discontinuities at the edges of the mirrors. A compact antenna range is supposed to mimic the far field of the test antenna in a much shorter distance. So, the running board on that van was a discontinuity, and the license plate was a plane reflector.

Personally, I don't want to be anywhere near an autonomous truck carrying 70,000 pounds plus of payload if the millimeter wave sensors can only see plane reflectors or discontinuities.
Tickerguy
Posts: 156031
Incept: 2007-06-26
A True American Patriot!
Report This As A Bad Post Add To Your Ignored User List
The point of adding such sensors is to augment the others. Humans can only see in the visual range, and our color sensitivity is uneven as well (which is why at night illumination "tilted" toward the yellow end of the spectrum produces better objective identification of objects at distance than light tilted toward the blue end.)

Machines can use sensors in multiple parts of the visual, IR and RF spectrum, all complementing each other. None is decisive on its own but they between them they can "see" at a range you cannot in difficult conditions. They may not be able to make positive identification at those ranges in poor weather but they don't have to -- as conditions deteriorate sufficiently they only need to be able to see well enough to be able to safely pull off and stop.

----------
Winding it down.
Flaps10
Posts: 6980
Incept: 2008-10-17

PNW
Report This As A Bad Post Add To Your Ignored User List
I feel a bit like a cowboy in the 1880s. Here comes progress
Login Register Top Blog Top Blog Topics FAQ
Showing Page 2 of 2  First12Last