avdmerwe

Ha that would be hilarious, but is unfortunately not true :)

There is a very good replica of the previous gen car in the Netherlands. I wouldn't enjoy driving it daily due to the noise https://www.youtube.com/watch?v=WVlgw90Ns7c&ab_channel=19Boz...

Of course, happy to chat. If you put my username in front of @signal.bio, you can email me

AFAIK the regs state that the signal for 'green' needs to have a human input, the wheel guns now mostly (video is from 2017) have torque sensors which inform a 'green' state when the human lets go of their switch.

Each team does things differently, but the front teams have a CAN network for pit equipment states to be shared for some form of automation.

(I currently work in F1)

This is triggered by sensors on wheel guns and the front jack. From the driver's perspective, this could even be sooner, as it'll take a while (~200ms) to react to the green light.

A lot of this is driven by the technical regulations, in order to keep some parity between teams with varying budgets. Most, if not all teams have as much software automation on the pitstop equipment (you can see the front jack release triggers the green light in the video!)

The nut is part of the wheel assembly. They have a fairly complex retaining system on each wheel hub, in case of a cross-threaded nut and unsafe car release.

Interesting to read this. It sounds like the opposite approach a rescue crew would take in a motorsport environment, in most cases.

I guess F1 is a much more controlled environment, due to car and circuit design working together as a safety system.

Perhaps F1 safety focuses on minimising injury when things go wrong, whereas aviation systems are optimised to avoid things going wrong in the first place?

Although I've spent the majority of my professional life driving around in circles, I have always been interested in technology, and worked in IT before I started racing professionally. Learning about materials, electronics, and manufacturing have been mainly through trial and error over the past 2 years, and working with some very talented people in their respective fields.

Great to hear that there's been some value in something I've said in an interview!

Exactly this. And I'd add that because we are now looking for tiny incremental improvements (most of the big, obvious things have been addressed already), we're going to need some data in the areas we're looking to improve.

Personally, I think it's really cool that anyone would post this on HN or even have questions about it!

Thanks! Strange to be posting, as I've been reading HN for years.

As with most things, this turned out to be more complex than we initially thought. The regulations in F1 meant that there were no off-the-shelf options for sensors. That, together with the fact that we needed to make things at low volumes, and potentially custom for drivers, meant that we had to develop the sensor ourselves. Fire rating and interaction with other safety systems on the car dictated a lot of our design choices.

The FIA has been instrumental in making this happen, as they are constantly trying to improve safety. Testing something like this at F1 races was challenging, as teams have less and less time to get their car development done. Everyone has been very supportive of the idea though, as teams always want more data!

Most drivers have done a decent amount of track time last year with various versions of the sensors, and we've chosen a design which ticks the most boxes. It is still very early days however, and I'm sure we'll have improved in almost all areas (comfort, performance & integration into clothing) by the end of the year.

The device we've developed can be used on many sites on the body, but for now the glove presents the 'easiest' to implement across the various clothing manufacturers. Further down the line, we'll probably fit more things in the earpieces, but this will require more than just anecdotal evidence as to what happens with the earpieces during large impacts. We want this data to be available in the most extreme cases. So although the ear would be a great place for SpO2 99% of the time when all is well, we don't know enough about how the sensor would interact with the sacrificial systems (helmet, head rest) in the accidents we'd most want this to work.

Bluetooth: I personally feel that the BLE standard is heading in the right direction for our use case, looking at the 5.0 standard and onwards. Lots of the current limitations will be solved by wider industry, instead of us having to do the legwork. Each car will have an on-car receiver which drops the data onto existing telemetry systems. In many cases, these systems are no longer functional after large accidents (antenna gets knocked off, car powered off etc), so we'd like to access the data with consumer devices in the worst cases. We already have iPads mounted in the car for other data, so BLE is the best candidate for now.

RF interference is actually also a reason we chose BLE: the 2.4GHz band is relatively quiet at race tracks, as most systems are in the higher bands these days.

I'd happily do an AMA, not sure how best to go about it though.

This will indeed not slow down anyone's initial response, but it will hopefully give us an insight into what happened before anyone gets there, or while the rescue evolves.

We would have really liked to know what happened while we were underway to the scene in this scenario: https://youtu.be/QOM5SzxEFGU?t=1m50s

And a driver under a few 100kg of Tecpro barriers takes a while to get to: https://www.youtube.com/watch?v=jJN-b-ovkQQ

Reading some of the press that this has had, it's unfortunate that some things have had a bit too much PR treatment. I'd also tend to disagree with 'immediately'.

We definitely don't expect this to be a silver bullet - it is the first (long overdue) foray into permanent physiological monitoring in F1, which will serve as a blueprint for further sensors and monitoring.

We chose SpO2 and motion as the first metrics, as they potentially offer us a lot of value with a single monitoring site.

This is meant to serve as an extra data point during the decision-making process when we first get to a driver, especially when we don't have immediate access to them.

As other posters have pointed out, we try to minimise additional risk to the driver and rescue workers as much as possible. Pulling a car out of the barriers quickly could potentially injure a driver that was perfectly fine.

The data will also be useful in post-analysis, as much of the time it can take up to 2 minutes for us to reach a driver at some circuits. If we can see that a driver wasn't moving and had low blood sats, that could be useful info when deciding on further treatment/imaging/etc.

I drive the F1 Medical Car, and came up with this system together with my colleague Dr. Ian Roberts, and have spent the past 2 years developing it. Any questions, ask!