A self-driving truck delivered butter from California to Pennsylvania in three days

[Antiwar]

At some point of automation im not sure its going to make sense to have capitalism anymore.

Say what? Automation is very capitalistic.

 

[bomberfox]

Say what? Automation is very capitalistic.

Ok? Does nothing to address my point.

 

[NatMorton]

A self-driving truck delivered butter from California to Pennsylvania in three days

Cupertino-based Plus.ai announces what’s believed to be an industry first: a cross-country commercial freight delivery by an autonomous truck.

www.denverpost.com

This sort of thing may be one of our ways out from the current issues with post COVID logistics (and associated inflation) issues.

I've thought for some time this is where self-driving vehicles can and will make (no pun intended) inroads. The driverless trucks can do the long hauls over the interstate system (roads that are simple to drive). If the destination is somewhere complicated (like, say, the City of Boston where many of the roads were laid out by horse and cart), then a driver can be waiting to take the truck through the last few tricky miles.

This could easily be safer than relying on a drowsy trucker who's been behind the wheel for the last 30 hours.

 

[PirateMk1]

lol, no. Where are we going to put it? How much is it going to cost? What would you do, if someone decided that the optimal route for this great new highway was right through your house?

What will happen is you'll have one driver baby-sitting the robots that are driving 5+ trucks in a platoon.

Meanwhile, most of the cars on the highway will gradually shift to automated driving. That will eventually make highway driving safer.

That's not going to fly. Not on the highways. The reason being simple if something goes wrong that requires actions for all vehicles in the platoon every other vehicle is on their own besides the one the human has control of. Platooning may be energy efficient but it is also dangerous and leaves little margin for action for the trucks behind the first one. If the first truck gets into an accident the subsequent ones will more likely be involved depending on the distances maintained. Platooning for fuel efficiency requires the vehicles to be directly in the draft of the proceeding vehicle the further back a vehicle is the less the effect of the draft and efficiency pickup.

Automated vehicles are going to run into the brickwall of liability. Who is, and in what circumstances.

 

[PirateMk1]

Autonomous vehicles are already operating on public roadways. Aircraft have the real potential to fall out of the sky and kill hundreds of people when something goes wrong; trucks don't. Commercial vehicles are regulated more than private vehicles but not as much as aircraft. I highly doubt that that's going to change much.

False. Commercial vehicles are one of the most regulated industries in the US more so than the air logistics industries.
Liability is going to be the driver of the change and what we see.

 

[eohrnberger]

Interesting article.
I realize the tech is improving.
- Still need human to fuel the truck
- Not sure the system could handle being directed by police to reroute do to an accident, road blockage, etc.

Fair.

Skynet is not yet ready for prime time.

Let's hope it never is.

 

[Antiwar]

False. Commercial vehicles are one of the most regulated industries in the US more so than the air logistics industries.
Liability is going to be the driver of the change and what we see.

I highly doubt that the physical components of trucks are going to change to aerospace quality control because of autonomous trucks.

Insurance covers liability. The carrier will be accountable just as they currently are, there just won't be a driver that's also blamed.

 

[PirateMk1]

I've thought for some time this is where self-driving vehicles can and will make (no pun intended) inroads. The driverless trucks can do the long hauls over the interstate system (roads that are simple to drive). If the destination is somewhere complicated (like, say, the City of Boston where many of the roads were laid out by horse and cart), then a driver can be waiting to take the truck through the last few tricky miles.

This could easily be safer than relying on a drowsy trucker who's been behind the wheel for the last 30 hours.

Regulations are 11 hours driving in a total of 14 hours continuous on duty. Once 14 hour clock starts it does not stop. A 10 hour minimum break is then required afterwards. OTR trucks are required to have an automated logging device which keeps track of driving time.

 

[eohrnberger]

That's not going to fly. Not on the highways. The reason being simple if something goes wrong that requires actions for all vehicles in the platoon every other vehicle is on their own besides the one the human has control of. Platooning may be energy efficient but it is also dangerous and leaves little margin for action for the trucks behind the first one. If the first truck gets into an accident the subsequent ones will more likely be involved depending on the distances maintained. Platooning for fuel efficiency requires the vehicles to be directly in the draft of the proceeding vehicle the further back a vehicle is the less the effect of the draft and efficiency pickup.

Automated vehicles are going to run into the brickwall of liability. Who is, and in what circumstances.

Assuming that the trucks who are 'platooning' are automated, and also assuming that platooning requires a feedback control loop between these trucks, it would seem to be logical to also assume that all the trucks in the platoon could be alerted as to what the first truck in the platoon 'sees' (perceives via sensors) and all the trucks would be able to take the needed corrective action in response.

 

[PirateMk1]

I highly doubt that the physical components of trucks are going to change to aerospace quality control because of autonomous trucks.

Insurance covers liability. The carrier will be accountable just as they currently are, there just won't be a driver that's also blamed.

Again wrong. Example carrier gets automated truck follows all recommended operating procedures to the letter, and additionally has the manufacturer DO all the required maintenance at the required times and ALL repairs. This presumes of course these vehicles are operated on nothing less than pristine highway. Something happens. Equipment malfunctions, what have you. Who is at fault? Who is at fault for a wrong decision made by the equipment that results in damage or injury or death?

Commercial drivers cover for a lot of manufactures substandard parts and construction on new builds and make older pieces of equipment outlast their design lives. If this is the case will the computers do so at first or will the build quality have to improve and well as replacement part quality?

 

[PirateMk1]

Assuming that the trucks who are 'platooning' are automated, and also assuming that platooning requires a feedback control loop between these trucks, it would seem to be logical to also assume that all the trucks in the platoon could be alerted as to what the first truck in the platoon 'sees' (perceives via sensors) and all the trucks would be able to take the needed corrective action in response.

True to an extent but also keep in mind that platooning is usually used for efficiency enhancement for fuel usage which requires the vehicles to be very close together less than a truck length. About 10 foot or less for maximum efficiency and it logarithmically goes down the further apart the vehicle gets. Eighty thousand pounds doesn't move quick no matter what the reaction times are. Safety would require a minimum distance of 2 to three truck lengths. At that spacing you lose the efficiency gains at efficient highway speeds. The platoon becomes an impediment to traffic especially going through cities at the the spacing of 0- to about 5+ truck lengths.

 

[eohrnberger]

True to an extent but also keep in mind that platooning is usually used for efficiency enhancement for fuel usage which requires the vehicles to be very close together less than a truck length. About 10 foot or less for maximum efficiency and it logarithmically goes down the further apart the vehicle gets.

Yes, and that's where the automation can probably perform better than humans, reaction times and faster information distribution between the trucks.

Eighty thousand pounds doesn't move quick no matter what the reaction times are.

Agreed.

Safety would require a minimum distance of 2 to three truck lengths.

Disagree. This is where the automated trucks following the lead truck in essence have the same 'visibility' of the lead truck, something that humans simply can't do. Think data net vs. a shout on the radio.

At that spacing you lose the efficiency gains at efficient highway speeds. The platoon becomes an impediment to traffic especially going through cities at the the spacing of 0- to about 5+ truck lengths.

 

[PirateMk1]

Yes, and that's where the automation can probably perform better than humans, reaction times and faster information distribution between the trucks.

Disagree. This is where the automated trucks following the lead truck in essence have the same 'visibility' of the lead truck, something that humans simply can't do. Think data net vs. a shout on the radio.

Reaction times, visability, and monitoring aside for the vehicles themselves they also have to interact with other vehicles and the environment around and have to be able to maneuver with that in mind. A bunch of vehicles running like a train will for the most part act like a train regardless of visibility reaction time and foreknowledge. It's one of the reasons convoying trucks was outlawed in most states. Trucks need much more maneuvering room than smaller vehicles because of the mass and mass profiles involved. Each truck regardless its position in the platoon has to be able to maneuver safely which requires space.
I am not against trucks becoming automated. I am actually for that. But right now and for the near future they are not viable. There are a lot of things that need working out.

 

[NatMorton]

Regulations are 11 hours driving in a total of 14 hours continuous on duty. Once 14 hour clock starts it does not stop. A 10 hour minimum break is then required afterwards. OTR trucks are required to have an automated logging device which keeps track of driving time.

And those regulations are never broken?

 

[eohrnberger]

Reaction times, visability, and monitoring aside for the vehicles themselves they also have to interact with other vehicles and the environment around and have to be able to maneuver with that in mind.

Yes, and the automation in doing all that is clearly getting very near to the 'breakout' point, where it is viable in the real world.

A bunch of vehicles running like a train will for the most part act like a train regardless of visibility reaction time and foreknowledge.

It is and it isn't. The trucks aren't on a rail without being able to maneuver, as a rail trail is.

It's one of the reasons convoying trucks was outlawed in most states. Trucks need much more maneuvering room than smaller vehicles because of the mass and mass profiles involved. Each truck regardless its position in the platoon has to be able to maneuver safely which requires space.

I suppose that depends on the situation. If there's a lane change possible to avoid a collision, all the trucks could follow the lead in doing so.

I am not against trucks becoming automated. I am actually for that. But right now and for the near future they are not viable. There are a lot of things that need working out.

I'm leaning to much the same conclusion, but those 'things that need working out' is dropping with every development.

 

[Visbek]

That's not going to fly. Not on the highways.

It's already happening. They've done live tests on highways for years.

FYI, highways are actually much easier for driverless vehicles, because they're much more predictable than city streets. You don't have to deal with intersections, pedestrians, bicycles, stop lights and so on.

The reason being simple if something goes wrong that requires actions for all vehicles in the platoon every other vehicle is on their own besides the one the human has control of.

Incorrect. They're going to coordinate whatever changes are required.

Platooning may be energy efficient but it is also dangerous and leaves little margin for action for the trucks behind the first one. If the first truck gets into an accident the subsequent ones will more likely be involved depending on the distances maintained.

Automated platoons will probably be about 50 feet apart.

Human truck convoys are typically around 250 feet apart.

There is no way that's going to make a difference at 60+ mph. The human can't even react fast enough for that to make a difference.

Then, you can add in the following:
• Robots don't get sick
• Robots don't get drunk or high
• Robots don't text while driving
• Robots don't need to take "go pills" to drive 24 hours straight
• Robots don't get distracted from the road
• Robots don't need vacation days
• Robots don't demand raises
• Robots don't need hotel rooms, food stops or sleeper cabs

Bottom line: Robots will be cheaper, safer, and more readily available than human drivers. While the tech isn't there yet, automation in the trucking industry is a done deal.

Automated vehicles are going to run into the brickwall of liability. Who is, and in what circumstances.

Hello? This is America. If there is an accident today, everyone gets sued. That won't change.

Sorting out the specific changes in liabilities will take time. However, that hasn't prevented any other industry from automating, and it won't stop automated cars and trucks either. The rewards are simply too great.

 

[PirateMk1]

And those regulations are never broken?

Much more difficult.

 

[bomberfox]

It's already happening. They've done live tests on highways for years.

FYI, highways are actually much easier for driverless vehicles, because they're much more predictable than city streets. You don't have to deal with intersections, pedestrians, bicycles, stop lights and so on.


Incorrect. They're going to coordinate whatever changes are required.


Automated platoons will probably be about 50 feet apart.

Human truck convoys are typically around 250 feet apart.

There is no way that's going to make a difference at 60+ mph. The human can't even react fast enough for that to make a difference.

Then, you can add in the following:
• Robots don't get sick
• Robots don't get drunk or high
• Robots don't text while driving
• Robots don't need to take "go pills" to drive 24 hours straight
• Robots don't get distracted from the road
• Robots don't need vacation days
• Robots don't demand raises
• Robots don't need hotel rooms, food stops or sleeper cabs

Bottom line: Robots will be cheaper, safer, and more readily available than human drivers. While the tech isn't there yet, automation in the trucking industry is a done deal.


Hello? This is America. If there is an accident today, everyone gets sued. That won't change.

Sorting out the specific changes in liabilities will take time. However, that hasn't prevented any other industry from automating, and it won't stop automated cars and trucks either. The rewards are simply too great.

The rewards for whom?

 

[Visbek]

The rewards for whom?

#1 is trucking companies
#2 is companies that need things shipped
#3 is consumers who get their stuff faster, and at a lower price
#4 is drivers, who will be safer

You may be tempted to say "truckers will get screwed." However, there's already a shortage of truckers, and the current drivers should be able to retrain fairly easily. I'm sure there will be lots of wringing of hands, but I don't think this will be anywhere near as bad as what happened to manufacturing.

 

[bomberfox]

#1 is trucking companies
#2 is companies that need things shipped
#3 is consumers who get their stuff faster, and at a lower price
#4 is drivers, who will be safer

You may be tempted to say "truckers will get screwed." However, there's already a shortage of truckers, and the current drivers should be able to retrain fairly easily. I'm sure there will be lots of wringing of hands, but I don't think this will be anywhere near as bad as what happened to manufacturing.

Retrain for what? You literally have to work for the companies who will own those trucks in order to even live so yeahhh i can see self driving trucks as a problem. Lots of people dont do it just for a hobby.

 

[EMNofSeattle]

lol, no. Where are we going to put it? How much is it going to cost? What would you do, if someone decided that the optimal route for this great new highway was right through your house?

What will happen is you'll have one driver baby-sitting the robots that are driving 5+ trucks in a platoon.

Meanwhile, most of the cars on the highway will gradually shift to automated driving. That will eventually make highway driving safer.

You won’t see even that.

Because a single failure on one platooned vehicle means cascading failures since the theory is supposed to be they travel drafting each other.

Automated trucking will never happen, so it will not make highways safer. To that end they can’t even make small cars function right for driverless operation.

 

[EMNofSeattle]

I highly doubt that the physical components of trucks are going to change to aerospace quality control because of autonomous trucks.

Insurance covers liability. The carrier will be accountable just as they currently are, there just won't be a driver that's also blamed.

It doesn’t matter whether or not physical components “change” what matters is if the truck is fully dependent on manufacturer installed systems you cannot just do your own maintenance on it.

If your LIDAR system for seeing obstructions needs a wiring harness replaced then a carrier that uses a 10 dollar wiring harness would have liability if the truck crashes. The response is the manufacturer will certify a specific wiring harness which you can only buy from them and of course being a monopoly it will charge a lot more money

 

[EMNofSeattle]

#1 is trucking companies
#2 is companies that need things shipped
#3 is consumers who get their stuff faster, and at a lower price
#4 is drivers, who will be safer

You may be tempted to say "truckers will get screwed." However, there's already a shortage of truckers, and the current drivers should be able to retrain fairly easily. I'm sure there will be lots of wringing of hands, but I don't think this will be anywhere near as bad as what happened to manufacturing.

The trucking companies wouldn’t really benefit. Because this type of technology (and it will never be ready for fully autonomous use anyway so this is only for discussion) would have to be leased.

And in any event drivers do much more then drive. If the technology only does 99% of the job you still need to hire a full time worker for the 1%.

In addition it would be expensive, likely more so then wages, which again, they have to pay anyway because most tasks drivers do are different then driving.

The robot truck isn’t going to be able to set up double trailer sets or break them down at different locations, if you’re driving construction trucks like cement mixers or side dumps, or dump trucks then the driver is actually doing the work with the product and driving is incidental. Even driving on the freeway there’s extensive work done in daily safety inspections, refueling, maintenance, keeping track of invoices, as well as actual security of the load. Can you think of an easier heist then ripping off a robotic truck? Using just two or three vans you can box it in and stop it and break open the container and take what you want, and on top of all that without a human driver it’s a mere property crime. Or if the truck has a systems failure and shuts down in Iowa somewhere people can break it open and rob it. Speaking of breakdowns who exactly is going to put the safety cones or flares out behind it while waiting for service?

 

[Antiwar]

#1 is trucking companies
#2 is companies that need things shipped
#3 is consumers who get their stuff faster, and at a lower price
#4 is drivers, who will be safer

You may be tempted to say "truckers will get screwed." However, there's already a shortage of truckers, and the current drivers should be able to retrain fairly easily. I'm sure there will be lots of wringing of hands, but I don't think this will be anywhere near as bad as what happened to manufacturing.

You think lots of people are going to retrain when jobs go away? Retrain for what? Concentration of wealth is going to continue. More people are going to become 'useless eaters.'

 

[Visbek]

Retrain for what?

Managing a platoon of self-driving trucks is likely to require a slightly different skill-set than driving a single truck. At a minimum, they'll need to understand how to interact with the self-driving software, and keep an eye on the coordination system.

You literally have to work for the companies who will own those trucks in order to even live so yeahhh i can see self driving trucks as a problem. Lots of people dont do it just for a hobby.

Only 10% of truck drivers today are owner-operators. New drivers usually have to pay for their own training.

Do you really think these are substantive objections?