While Melbourne had Formula One fever earlier this month, Ricciardo, Hamilton and F1 drivers in general are effectively on the endangered list. Data scientists will be soon taking the place of Formula One drivers. Not to physically sit in the car though. That’s the job of their artificial intelligence (AI) algorithms and software.
Welcome to Roborace, an extension of the Formula E electric car series. As the name suggests this is the driverless motor sports series. And it is, in so many ways, ground breaking.
The concept is a racing competition in the style of Formula E with a fleet of identical driverless cars designed by futuristic automotive designer, Daniel Simon. Hardware includes cameras, radars, and a Nvidia Drive PX2 “brain” sitting alongside the electric motors and ‘regular’ formula E mechanics. The software, however, is an open AI platform. The intention is a race with a field of the identical Robocars with separate data analytics, data science and engineering teams determining the AI algorithms for each car.
There are a few things making this so very, very cool:
- The cars are designed by Simon Daniels (who designed the Tron Legacy vehicles among others) and are truly striking – the very definition of cool;
- Identical cars and the open AI platform creates a level playing field;
- The cars travel at speeds up to 320kmh;
- By experimenting with collision avoidance algorithms in extreme racing conditions, we’re likely to improve in-car AI for regular cars faster;
- We get to watch a heap of firsts - like the first time two AI cars are on track simultaneously (the devbot cars – the proof of concept cars - did on Feb 2017), the first racing AI overtaking manoeuvre, the first full field of Robocars, the first crash (and intensive learning moment) and the first completed race without a collision; and
- It brings together teams of people around AI and the possibilities of AI.
- It’s the final point that locks in "cool".
- Just like in the business world, a lone amazing data scientist can make an impact to the business but there’s a physical limit. With a team of analytics professionals working collaboratively – data analysts, data engineers, data wranglers, visualisation experts and data scientists – the impact is more profound and sustainable to the business. These same kinds of people (in addition to the mechanical / engineering people) will be needed for each Roborace car to get the best outcome.
But every business doesn’t need a Roborace team to know if their analytics professionals have the right skills. You just need to join the IAPA 7500 people-strong analytics community as a member and take an IAPA industry-recognised analytics credential. Provided in partnership with Deakin Digital, these are the first evidence-based analytics credentials in Australia, benchmarked against the Australian Qualifications Framework so they are recognised around the world.
In the schoolyard of my farming childhood there were always bold “my Dad’s tractor is better than your Dad’s tractor” claims.
I can’t wait for Roborace-inspired future schoolyard claims to be, “my Mum’s algorithm is better than your Dad’s algorithm”!
Analytics professionals, start your engines!
Originally published on Linkedin Pulse
Watch the Roborace progress by following the “Inside Roborace” documentary series.
For those keen for the Robocar stats:
Motor 4 x 300kW
Materials Predominantly carbon fibre
Speed Over 320kph
Technologies 5 lidars, 2 radars, 18 ultrasonic sensors, 2 optical speed sensors, 6 AI cameras, GNSS positioning
AI power Nvidia’s Drive PX2 brain, capable of up to 24 trillion A.I. operations per second