Autonomous Vehicles

b0rn2sk8 said:
I think it's really odd that everyone is talking about cars but no one is talking about lings like trains, trams and even buses.

We have the technology right now to fully automate these systems right now without even changing the infrastructure (signals etc). All existing trains could easily be retrofitted with all the tech and sensors needed to run.

I know the drivers unions will very much grind it out to the bitter end resisting this but there have been far too many avoidable fatal train crashes in recent years in the 'west' and they were all human error (driver or signaller). I can't really see an argument against it and the savings for someone like TFL would be astronomical.
Click to expand...

The U.S. has begun to address this issue, calling for public comment.

https://uk.reuters.com/article/us-a...-self-driving-trucks-and-trains-idUKKCN1GD6C7
 
b0rn2sk8 said:
I think it's really odd that everyone is talking about cars but no one is talking about lings like trains, trams and even buses.

We have the technology right now to fully automate these systems right now without even changing the infrastructure (signals etc). All existing trains could easily be retrofitted with all the tech and sensors needed to run.

I know the drivers unions will very much grind it out to the bitter end resisting this but there have been far too many avoidable fatal train crashes in recent years in the 'west' and they were all human error (driver or signaller). I can't really see an argument against it and the savings for someone like TFL would be astronomical.
Click to expand...


DLR in London is driverless.

Most of the London underground is fully automated, the driver is just there in case (and because Unions say they have to be) and does not actually do a lot, as are many of the mainline systems in the UK.

Kobe in Japan has a fully driverless subway system, as does Lille in France, and Sao Paolo in Brazil.

Quite a few airport mass transit trains in the USA are driverless.



These things are on their way just as they are with cars, it is just the human bias against the mass integration of these systems that is stopping it really.

Just need a government with the balls to make these things law.
 
According to the Guardian today, there is new laser based system that will allow autonomous vehicles to see around corners---effectively "seeing objects before they come into view".

Thinking about the implications of this technology, if it can be brought out cost effectively, it would raise the safety record of an AV many fold, in my view. If the system could "see" that there is no way a vehicle approaching you from a blind spot today can slow down in time to avoid hitting you as you cross an intersection, the AV could know to stop or slow down to avoid the collision. There would be no way a conventional vehicle could ever claim to be as safe to drive as an AV.

Any views?

https://www.theguardian.com/technol...g-cars-may-soon-be-able-to-see-around-corners
 
Thinking some more about autonomous trucks and Waymo's announcement to haul equipment and parts from Atlanta to its data centres nearby, and considering the amount of capital now being invested by a number of start-ups in autonomous trucking as this article from Ars discusses (Uber, Starsky and Embarc to name but three), it appears likely that AV/trucks are a lot closer to commercial reality than most believe.

These companies surely must see a 5 year payback period for their investments.

What about other functions of the current truck driver likely stopping and delivering packages? This area is also getting the attention of the drone makers. In the US, the commercial drone industry wants to create a privately funded and operated air-traffic control network, separate from the current Federal system, to enable widespread operations at low altitudes. Companies like Google, Amazon, Boeing, etc see automated mobile and web apps being used to track and prevent collisions among swarms of small drones. In effect, they are looking to develop a private air traffic and control for drones and are conducting tests with NASA.

https://arstechnica.com/tech-policy...eyond-driverless-taxis-with-trucking-program/
 
Last edited:
Early Rider programme now officially launched in Phoenix, Arizona.

Some people find the experience equal parts thrilling and equal parts boring....it is simply that good at what it does.

At first the passengers giggle at the fact there is no driver in the front seat and then quickly tune out---one actually fell asleep.

Here is the 40 second video:

https://www.youtube.com/watch?v=QqRMTWqhwzM
 
McKinsey & Co, Consultants: "In less than a decade, 80% of all shopping will be delivered autonomously." Think about it. We are about to transform the package delivery industry as we know it. A robot that works 24/7 will cut delivery time from 2 days to 2 hours. Many believe that autonomously delivering packages will come well before robo-taxis for humans.

http://www.ttnews.com/articles/robot-rides-are-going-deliver-pizza-and-parcels-people
 
b0rn2sk8 said:
The main issue with deliveries is the last few meters from curb to doorstep/safe place.
Click to expand...

Yes. From the article above:

"Here’s how short-haul robot delivery might work, according to those creating it: You order an item online or on your phone and specify a window of time for it to be delivered to your home or office. As the package nears its destination, you get an alert telling you to come to the curb to retrieve your merchandise from a locked compartment in the vehicle. You unlock it with a code punched into a touchpad on the side of the vehicle or by scanning a barcode texted to your phone."

And further along, expect sidewalk robots and/or drones after that, to address the "50 foot problem."
 
Some of the comments from the Waymo CEO at SWSX conference recently in Austin, Texas:

1. Waymo cars are not connected to other Waymo cars (all the computing power is on board: for braking, steering, etc---but they share info with each other Waymo AV). They do not need a 5G wireless link to go anywhere. The car "carries" everything it needs to drive.
2. The car can see three football field ahead so it won't run into the dilemma of having to "choose" the "best" target to hit....the classic question that used to be tossed around.
3. By 2020, thousands of Waymo Pacifica SUVs (will have power sliding doors) will be offering ride sharing taxi services as well as many other Waymo "drivers" that are third party services (think Lyft, Uber, etc). You will use a Waymo app to call the service. There will be many form factors apart from the Pacifica SUV.
4. Waymo plans to have service in every major metropolitan city in the US by 2028 and in many cities around the world. Starting in Phoenix, Arizona and the sunny climate Waymo will move to "more difficult cities"---ie, where weather becomes more of a factor.
5. Currently it costs approximately $2.00 per mile to hire a Lyft or Uber and Waymo expects to price its service at approximately the same or a little less. Keep in mind that the average US driver spends about 20% of the time driving finding a parking spot. That time waste disappears with a Waymo AV.
6. Waymo expects to be "driving" a significant enough amount ---even 1% is significant---of the 10 trillion miles that humans drive globally every year (of which 3 trillion miles is US miles) by 2028.
7. Already more than 50% of Americans say they would be open to the idea of using an AV to get around.
 
Thinking some more about the Waymo CEO comments, these have the appearance of a Company (Waymo) that seems likely to become a public company in its own right in the near future. It may be that Alphabet/Google, Waymo's parent company today, may find it more appealing to separate Waymo from Alphabet due to risk mitigation....ie, the risks associated with the liabilities of offering Waymo ride-hailing taxis may be too great for the parent company to assume.

We could see this corporate "spin-out" occur before the end of this year. Expect a heavy ad campaign for Waymo globally to follow such a decision.
 
According to a recent NY Times article, there is a lot of discussion among tech leaders working on AVs about how to "drive" the AV in "edge cases" where the AV has difficulty figuring out a problem it encounters on the road.

Some companies are teaching their cars to drive themselves in all situations while some are working on remote control systems, also called teleoperations. Keep in mind that starting next month, California will allow the operation of AVs without a safety driver as long as the car can be remotely operated.

A remote operator could use the cars sensors and cameras to remotely navigate the AV out of trouble. One company uses remote control sensors to operate the AV is such edge cases, much like the technology NASA uses to navigate rovers on Mars.

How is Waymo handling the edge case? If a Waymo vehicle becomes confused — by, say, a new set of cones or a police barricade in the road — it can request confirmation from a remote human specialist. Once it receives confirmation of what it is sensing, the car — not the remote operator — then decides how to proceed.

The Waymo approach ensures that latency — a delay in the communications traffic — doesn’t compromise the car’s driving behavior by leaving a remote operator unable to react in real time.

Criticism of using a remote operator, however, is based on limited availability of wireless networks to the time it takes for remote operators to become “situationally aware” — informed not just of the immediate problem, but important context including the location or weather or any other variable that someone in the driver’s seat would already know.

So will AVs ever become truly autonomous?
 
Politics is still holding back the advance of AVs in the US. GM would like to start mass producing Cruise AVs next year, NHTSA's rules limit their ability currently to do so. Waymo has circumvented this problem by having its AVs comply with regulations of requiring a steering wheel and brakes but simply not allowing the passenger to access these features.

While this circumvents the restrictions, it means that Waymo car's have a drivers seat which is unusable and takes up unnecessary space.

Will the Senate reach a compromise? Safety requires them to do so.

https://arstechnica.com/cars/2018/0...-of-thousands-of-cars-with-no-steering-wheel/
 
You must log in or register to reply here.
Back
Top Bottom