Mobility Matters Daily #264 – TfL, AI, and the voice

Good day my good friend.

In case you wanted something to look forward to this September, Mobility Camp is back. We haven’t confirmed a date or venue, but we are looking out for people to help organise the event. You’ll get the chance to meet and work with some great people, and help to run an amazing event. If that sounds interesting, just email me. You can do the same if you have any suggestions for interesting news items or bits of research to include in this newsletter.

James

Thank you for reading Mobility Matters. This post is public so feel free to share it.

Transport for London’s finances – grow up, sort it out, and stop the theatre

Time to cut the c**p. Transport for London’s (TfL) funding is about to run out again, with the Department for Transport seemingly not happy with the pound of flesh that it has taken out of TfL in exchange for money. Because apparently keeping the transport network of your capital city running is some sort of favour. This has been a permanent debate for the last two years.

Let’s be clear here: this isn’t about public sector finances or TfL reforms. If it was, this would have been sorted a year ago. Transport is a political game, but the politics here benefit no-one. The longer this political game goes on, the longer investment and maintenance is put off, and the harder any reform becomes. So, political leaders. Sit down, shut the f**k up, hammer out a realistic deal, and think about Londoners for once.

London Underground train departs West Brompton station. there are several people waiting on each platform, with the train moving off away from the camera. It is passing under a bridge. There are flowers on the station platform

How artificial intelligence could help anticipate collisions (but beware of possible feedback loops)

An early release research paper gives some really useful insight into how artificial intelligence could be used to predict traffic collisions. To explain the system in plain English, the system tracks the eyes of the driver, compares it to dashcam footage, and predicts the likelihood of a collision. It successfully manages to detect collisions up to 4 seconds before i happens. And every second counts.

AI can also be used to map the likelihood of collisions and help city planners design better streets. It is clearly a potentially useful tool, and should be adopted for reducing road safety. But it won’t eliminate road collisions, just change the types of collisions that will happen. To what, who knows. But if we can save lives, we should look to do that.

How the voice affects how we react to emergency situations

One of the most tragic, yet fascinating, collisions in transport history is the Überlingen mid-air collision in 2002. Part of the reason why it is fascinating to me is the role of the human voice in the collision. In this instance, the pilots of Flight 611, having started to descend after heeding the warnings of their Traffic Collision Avoidance System (TCAS) heard a contrary instruction from the Air Traffic Controller, and climbed into the path of Flight 2937. The theory went that the pilots trusted the urgent call of the controller over the computer voice.

Some recent research touched on a similar point, by testing different in-vehicle voices to change driver behaviour awareness in automated vehicles. Simply put, the more submissive the language used, the more likely the people in the vehicle are to trust the system, and the more likely the driver is to regulate their emotions in a situation. Though this varies by the situation. The voice is very powerful indeed.

Random things

These links are meant to make you think about the things that affect our world in transport, and not just think about transport itself. I hope they do just that.