Beoz

Absolutely. The big shops need to sell hundreds of thousands of units before they even break even, and that's with years of R&D on tried and test tech.

Elon can't even get a few to market.

I think you will be soon finding a few investors cutting Elon loose.

Shard

https://www.theguardian.com/world/20...ple-to-retrain

I think Phil is being a bit ambitious here by predicting 2021 for widespread driverless, but maybe by 2031. Just retrain is good advice for the million or so UK drivers, a bit short on detail though.

GarryP

Not sure he's saying widespread autonomous, but 2021 is the date that most car companies have been targeting for the introduction of driverless. As I've said, I think the uptake from a working autonomous solution from there will be swift - because the push will be for the tech to be on as many cars as possible to recoup R&D, and because those that are on the roads a lot will gain more benefit, and will change cars sooner. By 2030 I expect the issue will be "how do we push these human drivers off the road?"

And that is why Tesla has such a high market cap, and why they will continue to find investment - they are pushing forward more aggressively into the areas which will define transport in the next decade. Model 3 is already set up with robotaxis in mind, they are that much ahead of the curve.

Beoz

This is why Tesla will see investors pull out. There's no point investing in a one trick pony when as you say, other car companies are doing the same.

The other car companies are not one trick pony's.

Shard

Yes, widespread was a bit of an embellishment, but it will be interesting to see how swift it will be. Such a cultural shift, and many people do actually enjoy driving. One parallel could be automatic versus manual, in theory manual should have faded away decades ago, but so many drivers prefer that extra degree of control. As you say, the car companies will be looking to recoup R&D and achieve economies. I guess it will be insurance rates and running costs that will push most humans off the road.

Shard

Many investors will be seeing Tesla as the frontrunner of what will be a new industry. A bit premature to anticipate their demise.

GarryP

As I've kind of pointed up before, there will be a collection of related factors pushing adoption

• First and foremost is the push towards robotaxis rather than ownership. Companies want this and for the city dweller and infrequent user it's likely to make a lot of sense. Hence why the realisation that the model 3 already has many of the attributes of a robotaxi as so interesting.
• Second, the high mileage road users are just the ones to change cars frequently and benefit from a vehicle that can drive itself (driving is a chore). You can see the repmobile being automated so they can do work on the go.
• Third, the hardware expected varied, but tends towards the cheaply mass producable. Most of the cost is in the R&D for the software, and thus there is a benefit on going mass adoption at sane price points early so you can optimise the income (particularly if you can sell subscriptions for updates).
• Forth, commuters hate the drive, the stop start. An autonomous vehicle for the commute that took the tedium out of it would have interest, particularly if you could tax deduct the company car.
• Put the above together and you might well find 10% of the vehicles on the road being autonomous within 3-5 years, but have them doing 30-50% of the road miles.
• And then there's the trucks, delivery vans, etc.

My guess is the S-Curve on the adoption is going to surprise people with it's speed. About the only fly is regulation and useless politicians clutching their pearls, however it looks like bribes are talking loudly and by the time the tech is ready, the laws will at least be half way there and not a major issue.

Stuck in Auckland

Enjoy the driving ? I think its another attribute they enjoy masked as "driving".

R&D which started 20 years ago is and has been recouped, through the higher models with the foundation technology already in place.

Manual to Automatic comparison? might as well be horse and cart to model T.

Look at adaption of smart phones

Beoz

Stuck in Auckland

I enjoy track days

I dont enjoy stop start motorway or inner city driving congested with pedistrians
.

Shard

Not sure if you can get Channel 4 programs down there in Oz, but there's a very interesting one on AI helpers that's just been aired. It's called "The Robot Will See You Now". Interesting reactions from the participants.

GarryP

Looks like if anything the timelines for autonomous vehicles is coming in, not stretching out.

GM lays out driverless car plans

They make clear that the target is robotaxies, mainly because they think they can coin it in

Stuck in Auckland

As an example in grocery warehouse automation fulfilment times are now down to 45 mins from order. being 15 mins sitting in order pool awaiting next 15 min wave. 10 min pick pack 5 min vehicle load. 15 min drive (warehouses situated near end customers).or click and collect 15 mins.

With less drivers going to bricks & mortar easier for more auto cars & trucks.

GarryP

Saw this article pointing out that, relatively quietly, things have moved along.

https://arstechnica.com/cars/2017/12...nyone-noticed/

Now google are testing autonomous vehicles WITHOUT safety drivers, on Phoenix streets. GM is thought to be not far behind.

Three years till 2021, and my guess is they will have commercial taxi services on city streets well before then - where the money is. Which is a pity since the long distance motorway cruising is going to be a key domain for acceptance I'd say. Melbourne to Sydney on full autonomous would be a game changer here.

ozzieeagle

Personally I think the first widespread acceptance will come via Trucking on the Nullabour between Ceduna and Norseman. That'll take away at least 5 driver shifts per 3 truck convoy, assuming the front truck is manned. Australia could possibly be one of the first places in the world where driverless is a viable proposition in trucking at least. 2020 I give it.