Ask HN: Software or Firmware Engineers in Formula 1 industry?
8 comments
F1 and automotive more broadly make use of all sorts of software to help their engineering.
The domains of Computer aided engineering are massive.
There’s FEA optimisation to design stiffer, lighter parts.
There’s CFD to design more aerodynamic surfaces optimising for a whole host of goals (drag reduction, downforce, air management, cooling)
There’s kinematic software to analyse the chassis-suspension system to understand behaviour in various scenarios (cornering/braking/accelerating)
There are multi disciplinary software that ties all this together (how does my suspension system in cornering impact my downforce and vice versus it’s a tightly coupled relationship)
The Non linear dynamics of suspension and tyres are pretty complicated (grad level mechanical engineering topics)
All of this is supported with modelling software, some of it third party, some of it homegrown.
All of the engineering simulation ultimately is piped into lap time simulation for a given track layout and variables can be tweaked to drive the optimal setting for that given track, car, driver, weather, etc.
The dance of F1 / race car design is given the regulations, time, budget, and other constraints how do you optimise for a host of non linear, chaotic, dynamic bits of physics, while making the car drivable for the specific driver.
Few people (Adrien Newey, Ross Brawn, et al.) can map out a vehicle concept completely, and then translate that into the efforts of 100’s of engineers to ultimately have a manufactured car ready to race.
The domains of Computer aided engineering are massive.
There’s FEA optimisation to design stiffer, lighter parts.
There’s CFD to design more aerodynamic surfaces optimising for a whole host of goals (drag reduction, downforce, air management, cooling)
There’s kinematic software to analyse the chassis-suspension system to understand behaviour in various scenarios (cornering/braking/accelerating)
There are multi disciplinary software that ties all this together (how does my suspension system in cornering impact my downforce and vice versus it’s a tightly coupled relationship)
The Non linear dynamics of suspension and tyres are pretty complicated (grad level mechanical engineering topics)
All of this is supported with modelling software, some of it third party, some of it homegrown.
All of the engineering simulation ultimately is piped into lap time simulation for a given track layout and variables can be tweaked to drive the optimal setting for that given track, car, driver, weather, etc.
The dance of F1 / race car design is given the regulations, time, budget, and other constraints how do you optimise for a host of non linear, chaotic, dynamic bits of physics, while making the car drivable for the specific driver.
Few people (Adrien Newey, Ross Brawn, et al.) can map out a vehicle concept completely, and then translate that into the efforts of 100’s of engineers to ultimately have a manufactured car ready to race.
Thanks for covering these aspects of different software in automotive industry. It makes me more curious as you mentioned the multi-disciplinary nature in the domain: How can they tune so many metrics and be able to optimize the whole car? Are racing events a huge part of the optimization journey, or small part?
You may be interested to know that Palantir works with Ferrari on exactly this kind of problem (sensors collecting data and making sense of it) - https://www.palantir.com/solutions/auto-racing/
Red Bull are a kdb shop: https://youtu.be/QxfdFWKo_pQ
I've seen Numpy code slightly blurred out of Mercedes videos on YouTube in the past.
I've seen Numpy code slightly blurred out of Mercedes videos on YouTube in the past.
Not at all exhaustive but the question reminded me of a couple of episodes of Talk Python To Me
- https://talkpython.fm/episodes/show/281/python-in-car-racing
- https://talkpython.fm/episodes/show/296/python-in-f1-racing
- https://talkpython.fm/episodes/show/281/python-in-car-racing
- https://talkpython.fm/episodes/show/296/python-in-f1-racing
It's all done in the UK. I expect it's paid just like the gaming industry (taking advantage of the passion of the employees for the sport)
Obviously there are mostly mechanical engineers, which is not my main interest.
All the teams are constantly collecting data from every part of the car during the races. There must be many sensors throughout the cars, right? They must need some software engineers to provide a good monitor/analyzer and embedded-system software engineers to do the firmware stuff.
Can any one share some F1 experiences?