[hawkinspeter]

From: https://nextcity.org/urbanist-news/bikes-not-self-driving-cars-are-the-technological-gateway-to-progress

[quote]It took a whole lot of noise from activists and campaigners for bicycles to be taken seriously at last year’s COP26 summit in Glasgow, and for active travel to be added to the declaration on accelerating the decarbonisation of road transport.

Beyond the serious lobbying from automotive industries, there seems to be a psychological block that prevents the bicycle from being accepted as a central technology when imagining the future of cities.[/quote]

Topic

[Bungle_52]

If there are any collisions

If there are any collisions though I would hope it would delay the introduction of autonomous vehicles, which could do a lot more damage, until they have had a couple of trouble free years but with the current government hell bent on deregulation I fear it won’t.

[chrisonabike]

I’m sure they’ve designed /

I’m sure they’ve designed / engineered in extra cuteness (maybe borrowing from Short Circuit / Wall-E?) for exactly that reason.  Although that may lead to them being stolen to become pets…

Presumably we didn’t go for a trained dogs / moneys on horseback solution because of cost concerns and fears they might unionise?

[mark1a]

Reminds me of the FBI mail

Reminds me of the FBI mail robot in “The Americans”

[hawkinspeter]

Bungle_52 wrote:

Looks like we’ll have some practical experience soon. DPD launching autonomous delivery in Milton Keynes. Currently learning routes according to this.

https://www.lapostegroupe.com/en/news/dpd-uk-to-launch-autonomous-delivery-robots-in-milton

A bit sceptical myself.

Those little robots are kinda cute.

I suspect they’ll have more difficulty from people going and kicking them over, than they’ll get from any collisions, but then they move slow enough and are small enough that collisions are unlikely to be serious.

[Bungle_52]

Looks like we’ll have some

Looks like we’ll have some practical experience soon. DPD launching autonomous delivery in Milton Keynes. Currently learning routes according to this.

https://www.lapostegroupe.com/en/news/dpd-uk-to-launch-autonomous-delivery-robots-in-milton

A bit sceptical myself.

 

[hawkinspeter]

I still think that

I still think that determining the class of moving object is less important than reliably detecting it. With unpredictable movers such as dogs and children, I’d say that the AI only needs to predict their current trajectory to determine if there’s any need to slow down or change course. The advantage of using AI is that it shouldn’t matter that it’s approaching a sheep lazily munching on the roadside – it should detect the proximity and lack of kerb and slow down appropriately to pass it and if the sheep suddenly makes a run for it, the AI would be able to detect that and react in milliseconds. It’s our slow reactions that mean that we have to have such good predictive models in our heads.

I’d consider that even with decent LIDAR, the AI models will still need to have object persistence and tracking, so if it detects a small child (or large squirrel) on the pavement and then it loses sight of it as it runs behind a parked car, it’d flag it as a possible hazard and reduce speed as necessary until it spots it again or passes the last known location.

[IanMSpencer]

My point is that to avoid an
My point is that to avoid an object you have to predict its path.

Consider an object detected and how a person deals with it if it appears to be a dog.

Is the dog accompanied? If yes, is it under control? If no, is it travelling in a calm predictable way? Is it actually a dog or is it a badger,a deer, a fox? Is the animal interacting with your presence?

Dealing with an animal is a peculiar experience. Think about sheep on moorland which may lie lazily as you skim their nose or bound across from a distance.

Then consider children and how they behave. (Think of the children!).

[hawkinspeter]

I think that Tesla is trying

I think that Tesla is trying to essentially use just cameras and AI to recognise and detect objects whereas other autonomous vehicles use LIDAR which is more expensive, but provides a 3d map of physical objects. I think it’s less important for a vehicle to distinguish between a dog and a pedestrian, but far more important to recognise that there is an object that must not be driven into.

[IanMSpencer]

I don’t think that the AI
I don’t think that the AI bods have really grasped that the problem is not just following road markings and looking for objects, but coping with imaginary road markings and invisible guides and conventions that have to be understood.

While computer programs can reliably follow asset of rules and don’t get distracted, the rule set required is massively more complex than the AI people hope.

Add in some shortcuts, so the likes of Tesla are trying to use sensor tech that doesn’t have the visual processing that humans have, where humans have glitches but generally are stunningly good at interpreting ambiguous images. As an example, try “Select Subject” AI processes in Photoshop-like programs. They take significant time when we immediately recognise the elements in the picture, but AI regularly add in extra odds and ends or misses an arm or a leg that a human would exclude or include because we can extrapolate body shapes.

We can then add in assessing reactions of pedestrians where we can recognise a pedestrian not paying attention or astray dog which requires an extra level of comprehension over simply plotting paths. This problem is exacerbated by an American view of traffic normally being isolated, physically and legally from pedestrians in many scenarios – interacting with a pedestrian environment isn’t being properly considered in many of the technology trials going on with the major American smart driving teams.

[brooksby]

I saw a billboard/bus-stop

I saw a billboard/bus-stop type advert this morning – cars / e-cars / bikes. 

(Emphasizing (I think) how congestion caused by e-cars looks suspiciously similar to congestion caused by ICE cars.)

(If I remember, I’ll try to get a photo at lunchtime).

[chrisonabike]

I suspect cycling will

I suspect cycling will continue to exist through most changes in the UK – but at a very low level.  After all, it’s still here despite the technological changes of the last over 100 years.  Equally all the so-call “encouragement” has done little to broaden the appeal.  (Possible exceptions – a minor increase in a few places like London).

On the other hand as long as you’ve still got some cycling and you’re prepared to spend a fraction of the road budget (say 10%) then cycling can become a mainstream form of transport – apparently anywhere.  Most people will simply choose the easiest form of transport whatever that is *.

I’m an optimist and there’s some hope in Scotland – we’ve managed to get a substantial chunk of transport budget for active travel.  So maybe a gradual change is possible.  For the rest of the UK I suspect it would take some drastic economic or political changes for mass cycling to happen e.g. a long-lasting fuel crisis, a revolution or some party like the Greens winning.  I can’t see any of the current main parties doing much different.

* Exceptions: if people already have cars and it’s convenient to drive, people drive.  If something feels unsafe or lacks cultural prestige or taboo that may override “ease of use”.

[Bmblbzzz]

I’m not sure the arrival of

I’m not sure the arrival of fully autonomous vehicles, whenever that is, will necessarily mean the decline of car ownership. People are  wedded to the idea of owning cars not only to drive them or (obvously) as status symbols, but as mobile personal spaces, a combination of living room and storage (and occasionally bedroom but that’s another matter). You can’t leave your possessions in the boot of a car that’s going to be hired out to a stranger, nor can you feel cosy-certain that the cup holders, music and air con will be as you like them. How do we got over this? I’m not sure; I expect there will be attempts at customisable bus seats, but that’s not a good way to go.

[hawkinspeter]

Found an article about a

Found an article about a Berlin startup (Swobbee) wanting to introduce a similar idea into Europe: https://techcrunch.com/2022/03/09/berlin-based-swobbee-wants-to-bring-micromobility-battery-swapping-to-europe/

They’re basing the business model on Gogoro in China: https://techcrunch.com/2021/10/10/gogoro-launches-battery-swapping-stations-in-china/

[chrisonabike]

I’m interested on your notes

I’m interested on your notes on e-taxis; for all i know this could be how it goes.  However I see a couple of points here which could mean this doesn’t become a major thing. (As you pointed out previously it does already exist).

Rich_cb wrote:

I see self driving cars as an ‘enabling technology’ that will allow cycling to flourish more quickly than it would in their absence. If we could eliminate all illegal driving behaviour our roads would be infinitely more pleasant to cycle on and more people would do so.

In my experience cycling with people who’re less confident about cycling (anecdata with a small sample size!) drivers not following the rules hasn’t been the biggie.  It’s just proximity to fast / large / noisy vehicles.  This seems to be bourne out in various studies / surveys ([1], [2], article about this).  Don’t know if all vehicles scrupulously followed the guidance (not all rules, see…) in the Highway Code that would have made a difference. 

Second – if owning a vehicle declines this may (subject to whatever else changes in the future) mean car clubs instead.  That can also be a good thing but doesn’t necessarily give as much “better driving” benefits.

I agree about virtuous circles though.  Although excellent provision for cycling is still not always “given” in e.g. The Netherlands apparently it’s now a less partisan issue where mistakes and oversights can generate a noticeable public response.

[hawkinspeter]

I’ve heard of courier

I’ve heard of courier companies (also in Asia – maybe India?) that send out riders on e-mopeds and do a similar swapping of batteries to minimise any down-time. That’s a great idea as the transit hub can have all the specialist hardware for charging the batteries.

[Author]

Replies