I think the interesting thing about Helbing's modelling is that it suggests that autonomous systems are sufficient to produce an emergent result that works whereas centralised control is much more difficult. An analogy might be with a flock of geese which can be modelled as flying slightly behind and to the right or left of the goose in front. That's the only rule needed to create the flock. They don't all have to have the programme of information of the lead goose.

The sudden and frequent speed limit changes on smart motorways do cause traffic jams rather than preventing them. I have seen 2 accidents on the M42 past Solihull caused by the sudden 40mph limit imposed through a junction and speedo watching not road watching. Road watching would be much safer.

Also, late signage and provision for exits. I recall the London bound exit from the M40 to the M25 used to be a single lane starting about a quarter of a mile before the peel off. Partly on the basis of some of Helbing's work I wrote to DoT and the Highways Agency and pointed out that this was causing major jams. Whether at my instigation I shall never know, but they made it into two dedicated lanes stretching back over 3/4 of a mile. I had found it much faster to go to the end of the M40 at Denham and come back to the junction from the opposite direction than wait in the exit queue.

Helbing also did interesting work on the optimisation of traffic flows in cities, looking at light phasing and operating rules, roundabouts and uncontrolled junctions..

It doesn't add up...,
Interesting that they describe a lithium mine as a potential

... revival of the county’s industrial past as a new green energy hub.

Since the lithium will have no part in generating "green" energy, but might be necessary to make that energy useful, you could equally describe a new copper mine in the same glowing "green" terms. And if your wind farm is backed up by a gas turbine, that gas turbine must be green too.

On ACC, I only skimmed it, but the paper seemed to be based on mathematical models and ideal behaviour. To be sure, ACC systems will do a better job of estimating speeds and of being attentive, but their failings in cognition can lead to unpredictable behaviour which might make things worse. As you say, more sophistication needed, especially when you're dealing with a mix of ACC and human decision makers.

There is probably more to be won at a higher level using GPS speed tracking and variable speed limits.

DaveS,
The China question is the litmus test: is their concern about the environment or is it politics? Latest at Jo Nova's suggests they haven't been achieving their environmental aims.

Jun 27, 2023 at 9:45 AM | tomo

Good to see that the responses, such as they are, take a realistic view of this body, probably not what harrabin was hoping for. With a Chinese bloke on the supervisory board I assume they will be focussing most of their efforts on the country that now produces the majority of the world's steel...

IIRC Dirk Helbig at ZTH did show that ACC had significant potential to help avoid motorway traffic jams that are caused by reaction braking in near tailgating conditions that can propagate into a jam backing up many miles.

https://www.researchgate.net/publication/2174945_Jam-Avoiding_Adaptive_Cruise_Control_ACC_and_its_Impact_on_Traffic_Dynamics

Requires some more sophistication though.

An update on United Downs

https://www.cityam.com/lithium-discovery-could-revive-cornwalls-industrial-past-as-new-green-energy-hub/

They got more funding. But 25MW doesn't sound like a very economic output to me on the energy side. They need the lithium to work.

Meanwhile Cornish Lithium is threatened with bankruptcy.

https://www.cityam.com/cornish-lithium-faces-uncertain-future-as-it-scrambles-for-10m-lifeline/

MikeHig,
Adaptive cruise control suffers from that problem (amongst others): it just latches onto the car in front and aims to maintain a fixed gap (in time or distance). The fully autonomous enthusiasts would argue that their vehicles *will be* better than a good driver. Like a chess computer, the central controller will look ahead at the spreading tree of possibilities of what it can do and what each identified actor might do. It would be reasonable to extend this tree at least as far as the time it would take to bring the vehicle to a stop. If the traffic further ahead is braking, the autonomous vehicle will ease off accordingly.

Which sounds great in theory, but it's at the level of Monty Python's how to play the flute: "blow in this end and move your fingers up and down on the keys". Not so simple when it comes down to brass tacks.

Ambitions did get far ahead of capabilities. That article is about 6 years old and included this remarkable sentence:

Volvo plans to release its first autonomous cars by 2020 and has pledged zero fatalities or serious injuries from all its cars by that time.

Maybe they expected to be out of business.