Self driving cars are going to upend this industry (eventually). Many mining companies are actively working on getting their mining equipment to be automated to do the riskier jobs without risking drivers. The same ideas apply here. Remove the driver and suddenly you have a truck that runs deliveries 24/7 and only stops to refuel and get serviced. One would hope that as the technology gets more sophisticated the number of accidents would drop, the vehicles could safely drive and react at higher speeds, and of course self optimize their routes and driving styles to save on fuel.
It's the loading docks, the fueling stations, the truck scales and all the rest of the infrastructure that supports trucking.
All of those are going to need to have "automated truck" support as well, not to mention, when an 18 wheeler blows out a tire, you need some skills to bring the truck to a successful stop, but even more skills and equipment to effect the repair.
Mining operations are easier because they work on their own property and aren't required to mix with other passenger vehicles.
I know you said "eventually," but I suspect that it's going to take at least 50 years before we're anywhere even close.
> All of those are going to need to have "automated truck" support as well
Do they?
The trucks don't need to automate 100% of tasks for initial implementation, the first ones will be driverless in-between cities (or driver-nap in-between cities) with manned fueling, loading docks, scales, etc.
This will require some changes, but I don't think it will be too much of an issue to move toward full-serve gasoline service in the interim.
If you want them to be economically competitive, then yes.
The largest current expense is the driver salary, mostly due to the low quality of the working conditions and extended stays away from your "home." If you're still carting people around in the truck, then you will have a difficult time competing with large personnel and technology expenses.
What I'm trying to suggest is that a fully automated nation-wide delivery service is going to be exceptionally difficult and will take a long time.
This is mostly from the perspective of OTR/long-haul transportation. You'll likely see faster adoption on smaller scales such as in-city delivery services like grocery delivery and the like. It will blossom outward from there but it's still going to be a very long process.
The idea that you can automate driving a car between cities seems feasible to you (with traffic, and variable weather), but automating fueling doesn't? Really which seems like a more complex task to you?
Loading and unloading the vehicle at docks will adjust to meet a happy medium. As you already mentioned the single biggest expense is the driver, eliminating that is key, everything else will adjust accordingly.
> when an 18 wheeler blows out a tire, you need some skills to bring the truck to a successful stop, but even more skills and equipment to effect the repair.
An automated truck won't need a conventional cab. It could essentially be a powered sled that rigidly attaches to a trailer with much less forward projection. This would be easier to control in emergency stops.
Except that trucks usually don't haul the same trailer day in and day out.
Also when a truck is loaded the driver has to check weight, balance and shift the pins around on the trailer and 5th. Who does that what the load shifts because the truck made an evasive maneuver out on the road in bfe middle of Utah? There's a lot of aspects to driving otr that people don't realize.
This industry already operates on tight margins so bringing in new tech which may be beneficially in the long run is hard to justify for most small companies and owner operator outfits.
you need some skills to bring the truck to a successful stop
True, but that's skill most drivers don't have. Add in-wheel pressure and temperature sensors-- plus scheduled maintenance checkups and that problem mostly goes away. There are also numerous examples of adaptive control systems landing aircraft after major failures (quadcopters with 3 props, planes with 1 wing) that humans can't handle.
You bring up good difficulties. Still-- most of those problems have solutions and I sure hope it's sooner than 50 years from now we see self-driving long-haul trucks on the road.
> True, but that's skill most drivers don't have. [...] after major failures (quadcopters with 3 props, planes with 1 wing) that humans can't handle.
These are decent examples, but don't most of them only account for the safeyy of the controlled vehicle itself? I mean, it's impressive that a quad copter can land itself after a partial failure, but can it do it over a crowded city? With a hazardous load? Can the control software understand NFPA 704 of it's own cargo, plus the cargo of other vehicles near it and make appropriate decisions? All of this possibly near the maximum weight ratings for it's transportation type?
I'm _all_ for automation, and I think it's the next most logical direction of our society; however, I think people tend to underestimate the totality of challenges involved especially for cargo transport.
One huge advantage of a self driving truck is it can easily wait 4+ hours for a tow truck to arrive for minor repairs at little cost. Or be swapped out for another rig and hauled off for more major repairs.
If your logistics operation allows for that, sure; most do not. You want to do as much work with as little miles as is possible, so turning the trucks over quickly is important.
There comes a point where industry overhead buffers you.
What I mean by that is if the most cost effective big rig truck gets say 10MPG then that cost gets built into the market rate. If a new truck design gets 15MPG and costs more you can figure out the costs vs what's out on the market and see if it's an improvement.
As to self-driving trucks, sure there are downsides. But, I suspect the net efficiency gains can offset them in most situations. Once most shipping companies switch the downsides will just be 'industry overhead' until their dealt with.
> What I mean by that is if the most cost effective big rig truck gets say 10MPG then that cost gets built into the market rate. If a new truck design gets 15MPG and costs more you can figure out the costs vs what's out on the market and see if it's an improvement.
I see your point, but that's a particularly poor example. Most shippers will pass the fuel costs as a "fuel surcharge" on to the customer. If you're an owner/operator, this is the most lucrative part of the business, and is really the whole reason to be an O&O in the first place. Trucking is just a really strange business.
As the largest factor in accidents other than DUI is inattentiveness[1](tired or texting) then self-driving cars with limited ability for drivers to perform actions that could cause an accident or effect efficiency would be a huge boon.
Imagine saying, "sorry my car cut off your car!" with sincerity and a little tongue-in-cheek humor as well.
Then there's also the efficiency factor as far as maximizing capacity in lanes and maintaining speed to streamline commuting. Merging could be vastly improved if cars were able to zipper merge better.
These self-driving vehicles will most likely have a lower base risk for accidents - they don't get tired, they don't get distracted.
However their potential for a Black Swan event is enormous. They could run perfectly for 5 years then an unexpected bug could cause 50% of vehicles to spontaneously accelerate to max speed and then pull a hard right..
We are still a long way from vehicles without human occupants.
The most modern lorries by Mercedes-Benz are very autonoumous, but still need a driver to be with the lorry.
Since lorries rarely just go from marked up parking spot to parking spot, but at their destinations probably have to navigate unconventional layouts, etc, they will for the forseeable future include drivers.
But what I expect to see happen, is that the logistics will become more decentralized, since the driver can now spend more time on administration, and less on driving, essentially taking over the job of the company command central.
The amount of information an automated vehicle would have to be able to decipher to be capable of being truly autonomous is pretty much pushing beyond the most advanced A.I. and visual learning academia has discovered today.
I don't know what gps resolves to but it's definitely not down to the foot that's required for some of the hairy situations truckers encounter on the road and our visual learning technology is terribly insufficient for even the most basic road conditions much less your average New England construction zone.
Even google cars can't handle weather or potholes or unpredictable situations well. From what I read they had to be coded for very specific paths around San Fran and any unpredictability would cause them trouble.
A more likely scenario in my mind is Drone trucks. Why can't a trucker head to the office drive his truck from his desk then go home at night?
Combine that with some gps autopilot and/or magnetic strips in the road as a guide and that's totally realistic even with todays technology.
These are more likely scenarios.
Edit: even cooler would be autonomous or drone blimps carrying cargo!
> I don't know what gps resolves to but it's definitely not down to the foot that's required for some of the hairy situations truckers encounter on the road
Precision positioning, using correction data similar to WAAS (http://en.wikipedia.org/wiki/Wide_Area_Augmentation_System) but heavily localized can provide sub-centimeter positioning data. This is already in use today for agriculture (self-driving combines, with waypoints configured from the farmer's desktop at home).
> A more likely scenario in my mind is Drone trucks. Why can't a trucker head to the office drive his truck from his desk then go home at night?
UAVs can stay flying if a signal is lost, or navigate using GPS to the nearest friendly runway to autoland if they lose touch with their pilot in the Las Vegas area (where most UAVs are piloted from). Ground vehicles traveling 60+ mph do not have this option. Therefore, AI is a superior option.
Drone trucks would also provide a metric fuckton of training data for autonomous vehicle research. I dunno how much effect the lag would have on safety.
I could say the same about mechanical or electronics of a vehicle
"I've always been leery of vehicles. They run on mechanical parts. Mechanical parts are only as good as the mechanical engineer that designed them and the quality of material. What if a newly introduced part was not well tested and developed a stress failure due to design or material deficiency. Be it in the middle of rush hour traffic, or hurling down the highway at 70mph? How do we account for the issue of buggy mechanical parts or electronics "
See how that works? Everything other than your own two legs is man made. It's just the level of maturity of the process and testing and time that eventually gives you enough confidence to start trusting any human made device, whether it is software, mechanical, electric, electronic etc...
> I've always be leery of self driving cars. They run on software. Software is only is good as the guy programming it.
Software is as good as the best team we can find to write it, and then it can get constantly improved - unlike meat drivers, whose firmware cannot be easily flashed. Moreover, when switching from human control to software you're trading the need for quick evaluation and reaction of driving conditions repeated times number of drivers for a long, concentrated time spent on development and testing. It seems to me like a good trade-off. You can spend all the time you want writing, testing and fixing software up until you reach the point it drives better than humans.
And then you have usual, obvious advantages - self-driving cars won't get tired, emotional, distracted by texting and definitely won't be driving in blatant disregard of traffic laws because "they're pointless and that's how real drivers do things".
You mean you'd prefer humans to do the driving? Youtube is full of human examples - start here as an example https://www.youtube.com/watch?v=qwL7S--E5oQ - and note that it ranges from minor fender benders through some very serious injuries and destruction.
There are lots of examples of not being able to stay in lane, pulling out, not seeing stopped traffic, not paying attention to surroundings. Humans can't even get the trivial simple things right in broad daylight.
(Yes I am aware that you are in the 97% of drivers who are absolutely convinced they are above average. That doesn't make it true. Yes the video mainly has clips from Russia - you can find USA based ones too. And a 9/11 happens on US roads each and every month. People are dying in large numbers because of human drivers.)
All the same, I have a large amount of trust in my own driving. I trust my driving more than my own programming. And I sure as hell trust my code more than some sketchy offshore team's. It's only a matter of time before the first car segfaults.
Plus it's a huge security risk. Cars are internet-connected these days.
Well, for one you can be pretty sure that self-driving car software isn't going to be written by hippie webdevs changing frameworks every other week. Companies and programmers that tend to work on this kind of software have pretty good track records - we don't see planes falling out of the sky or medical machines randomly killing patients all the time. Uou're going to have an app store for your car drivers.
Cars have it simpler than planes etc too. In particular they can always fail safe by stopping - something that isn't applicable to planes. Other complexities aren't present either due to the 2d nature of roads - for example car software doesn't have to worry about stalls, coffin corner or similar.
I believe that self driving cars will primarily be owned by companies with fleets of them, rather than individual ownership we have today. It doesn't make sense to have them sitting idle 23 hours a day. The owner companies will have a very strong incentive for safety since even rumours would be bad for business. They will also be in a better position to do something about it due to having multiple vehicles.
> It doesn't make sense to have them sitting idle 23 hours a day.
Well, I agree. That's why I expect that, along human drivers, the very concept of car ownership will disappear. It's much more efficient to have those cars available on call; they can serve someone else while you're not using them, and they can be parked&recharged at night with much better space-efficiency than cars are parked today.
Especially on a highway. I am talking about a situation where normal operation isn't possible such as due to hardware or software failure. Of course it isn't ideal, but wouldn't be much of a problem if the other vehicles on the road are driverless. (To be clear I mean they can just stop in the middle of the road if things are bad.)
A plane can't just let go of all control surfaces. Heck it can't even leave them where they are. It must do something and always must provide some degree of control. A flying plane is moving at several hundred miles an hour and cannot stop.
> Software is only is good as the guy programming it.
I guess chess computers can never beat a human then. Because the software is only as good as the guy writing it.
> What happens when the software develops a bug? Be it in the middle of rush hour traffic, or hurling down the highway at 70mph?
What happens when a human is inexperienced/texting/drunk/sleepy? Humans have plenty of "bugs" but you don't notice them.
> For all of the great things we know self driving cars can potentially bring to the table, how do we account for the issue of buggy software?
Trials. Testing. Every accident (which is inevitable) will be scrutinized and used to improve the system (unlike the 40k annual deaths in the US where that unfortunate experience helps nobody, except perhaps installing a traffic light or barrier).
Back when I wrote video games, I often had to explain to people that was a terrible player of these games. Many people have a default implicit assumption that you can't make something that you are not a master of. Similarly, it's hard to appreciate how software can encode the accumulated "smarts" of literally thousands of contributors, all running at inhumanly fast speed.
Cars currently run on millions of lines of software.
Airlines run on even more.
Medical equipment is heavily run by software.
Software already runs our lives, and worrying about the quality of the underlying software is a strawman. When was the last time a code bug took out a car or an airliner? Some industries have very stringent requirements and QA processes to make sure the code is bug free and can't kill people.
"The crash of Japan Airlines Flight 123 on August 12, 1985 is the single-aircraft disaster with the highest number of fatalities: 520 died on board a Boeing 747. The aircraft suffered an explosive decompression from an incorrectly repaired aft pressure bulkhead, which failed in mid flight, destroying most of its vertical stabilizer and severing all of the hydraulic lines, making the 747 virtually uncontrollable."
"On March 3, 1974, Turkish Airlines Flight 981, a McDonnell Douglas DC-10, crashed in a forest northeast of Paris, France. The London-bound plane crashed shortly after taking off from Orly airport; all 346 people on board died. It was later determined that the cargo door detached, which caused an explosive decompression; this caused the floor just above to collapse."
"On February 19, 2003, an Iranian military Ilyushin Il-76 crashed in mountainous terrain near Kerman in Iran. The official report says bad weather brought the aircraft down; high winds and fog were present at the time of the crash.[21]"
"On May 25, 1979, American Airlines Flight 191, following improper maintenance and the loss of an engine, a McDonnell Douglas DC-10-10, lost control and crashed near O'Hare International Airport in Des Plaines, Illinois. The crash resulted in the deaths of all 271 passengers and crew on board, as well as two people on the ground."
"On July 11, 1991, Nigeria Airways Flight 2120, a Douglas DC-8-61 aircraft operated by Nationair Canada, crashed in Jeddah, Saudi Arabia after two tires ignited upon takeoff, leading to an in-flight fire. All 261 people died."
None of the examples you've cited are software bugs, all physical failures or pilot error.
If you're looking for something that could be software related try Air France Flight 447, though even that is ultimately pilot error. You could also use Ariane 5 flight 1, where a software problem caused vehicle failure.
At least the bug is able to be diagnosed and traced along where it started from all the way until the accident where it caused a problem.
Try diagnosing someone suddenly getting pissed off and cutting off another driver which sets off a chain reaction of the events leading up to a multi-vehicle collision. The driver may attempt to cover-up, deny or lie about what happened to protect themselves whereas the software would not have such self-preservation and would report things as they occurred without being concerned with assigning blame or establishing guilt.
Software is not going to turn its head and rubberneck at an accident and in turn, rear end the car in front of it.
No, they're warning us about an AI which could outsmart or outperform us, and lead to our demise as it pursues the goals we've encoded into it without realizing they lead to a result we don't want.
I don't think so. I think a case can be made that he's talking less about SkyNet and more about the risks something like an automated trading algorithm (or planning system, or any case where we put too much trust in a system as infallible) ten years in the future causing major disruption due to too much trust in a fallible system as being infallible.
There's also the possibility that a true AI (not just a complex expert system as above) may be fallible just like a person in not being able to correctly account for the unknown unknowns.
In the end I think it's about the difference between automating positions to make them more efficient (meaning it's still possible to do manually but less efficiently) and relying on things that are essentially impossible to do manually, such as an economic planning AI which may outperform regular economists 95% of the time. The question is what happens in that 5%, and will it catch the problem, and if not, will we keep listening to it as it takes us down a dangerous path.
This guy is fantastic for being so proactive in his approach while being careful to (what appears to) run the system in such a way that it does not interfere with the way the truck operates. Not to mention the saved seconds or minutes which add up after a while.
Kind of fascinating that his preferred interface is physical buttons built into the cab, and a barcode scanner.
I can't imagine no one has built mobile software in this space yet, so there's got to be a reason he doesn't have a tablet sitting on his dashboard. He even built an app but curiously didn't opt to use that as his primary interface for the other functions. Says something about what a working truck driver actually needs & wants vs what a software developer would expect as the solution here.
I just realized that truck drivers may also be regulated regarding what types of devices they can have when driving, which would explain a lot. Yellow cabs in NYC for instance are forbidden to use GPS due to safety reasons and generally rely on dispatchers to get around. Truckers probably have even more safety rules to follow
In the US, I do believe that this technology exists. Whether or not all shipping firms have adopted it, I'm not certain.
As for the UK, I have even less of an idea. Either way, this fella had to call in each time a haul was transferred, so it seems that he's saved himself and the company a bit of trouble.
It does exist in the US. The ones I've seen are linked to GPS trackers that are added to fleets, so it's more a company adopted thing than a driver adopted thing.
The amount of data that you can get is pretty impressive from tracking a driver in real-time. There are even versions with cameras pointed at the driver and the road to record any "adverse events". In this way, it's less focused on the driver and more focused on fleet management.
I interviewed for a job in Las Vegas last year where I was shown this sort of technology for the trucking/shipping industry, albeit more server-based and not so distributed.
likely in different forms, think UPS/Fed Ex with their tablets and GPS connected trucks. Many trucking companies know exactly where their trucks are, they don't need to be told when they left; likely only why they haven't
Instead of going through all the hassle of doing barcodes and getting a barcode scanner, he could make a mobile application with buttons for each predefined message to send over email/twitter. Probably something like this already exists.
they appear to always be on gpio 23 and 24, which a comment implies is for making beeping sounds (I'm guesing two-tone?). I'm guessing it's just to play beep sequences after it has done X or Y.
