Student Teams Master Energy Efficiency Strategies on Circuit Paul Armagnac, Nogaro 2024, empowered by a Data & Telemetry Bootcamp hosted by Schmid Elektronik.
𝗕𝗼𝗼𝘁𝗰𝗮𝗺𝗽 𝗣𝗮𝗿𝘁 𝟭:
𝗖𝗿𝗮𝗳𝘁 𝘆𝗼𝘂𝗿 𝘄𝗶𝗻𝗻𝗶𝗻𝗴 𝗲𝗱𝗴𝗲 𝘄𝗶𝘁𝗵 𝘀𝗶𝗺𝗽𝗹𝗲 𝘀𝘁𝗿𝗮𝘁𝗲𝗴𝗶𝗲𝘀 𝗳𝗼𝗿 𝗶𝗻𝘀𝘁𝗮𝗻𝘁 𝗶𝗺𝗽𝗮𝗰𝘁!
We first reviewed the learnings from the previous webinar series of three hosted by Schmid in march 2024. Then we figured out how to apply them on the real thing: the race track!
With data dashboarding combined with Minimum Viable Products (#MVP), students where guided through iterative learning for immediate performance gains.
They learned how to familiarize themselves with a track and how to come up with several race strategies that they think will work best. This included meaningful and executable burn & coast patterns for their driver to memorize. Ready to be tested on track!
𝗕𝗼𝗼𝘁𝗰𝗮𝗺𝗽 𝗣𝗮𝗿𝘁 𝟮:
𝗔𝗽𝗽𝗹𝘆 𝘁𝗵𝗲 𝗦𝘁𝗿𝗮𝘁𝗲𝗴𝘆 𝗽𝗵𝘆𝘀𝗶𝗰𝗮𝗹𝗹𝘆 𝗼𝗻 𝘁𝗵𝗲 𝘁𝗿𝗮𝗰𝗸 𝗮𝗻𝗱 𝗖𝗿𝗲𝗮𝘁𝗲 𝗬𝗼𝘂𝗿 𝗗𝗮𝘁𝗮
While moving along the straights, climbing the slopes and cutting the corners, the drivers focussed on consistency in the driving patterns. The goal: coming up with precise speed profiles and energy consumption traces.
In the transition from theory to practice, they honed several racing strategies and tactics that were precisely tailored to the respective vehicle, driver and track. And improved step by step. Not based on complex computer models, but with iterative learning.
𝗕𝗼𝗼𝘁𝗰𝗮𝗺𝗽 𝗣𝗮𝗿𝘁 𝟯:
𝗧𝘂𝗿𝗻𝗶𝗻𝗴 𝗥𝗮𝗰𝗲 𝗧𝗵𝗲𝗼𝗿𝘆 𝗶𝗻𝘁𝗼 𝗮 𝗵𝗼𝗹𝗶𝘀𝘁𝗶𝗰 𝗦𝘁𝗿𝗮𝘁𝗲𝗴𝘆 𝗮𝗶𝗺𝗶𝗻𝗴 𝗮𝘁 𝗯𝗲𝘀𝘁 𝗧𝗿𝗮𝗰𝗸 𝗣𝗲𝗿𝗳𝗼𝗿𝗺𝗮𝗻𝗰𝗲
Through iterative cycles over two race days, teams carefully crafted data through telemetry, ensuring consistency in speed and energy profiles. As their MVP evolved, they developed a deep understanding of diverse racing strategies and driving tactics, mastering the art of delivering precision-driven, high-quality attempts.
In this third boocamp part, we looked into several consistent and inconsistent race data profiles and how to turn data into knowledge.
Next, students learned how to load their race data into the Jupyter Lab to apply individual analytics with the #Python programming language.
Finally, they learned how to solve the famous multigoal challenge in #ShellEcomarathon racing: maintain safety and improve energy and time efficiency from the start to the finish line, covering straights, slopes and corners.
𝗧𝗵𝗲 𝘁𝗿𝗶𝗰𝗸: describe the race scenario with the #nodes and #edges of a mathematical network #graph and apply a graph transformator like the #Dijkstra or #A* to reach a global optimum.
The goal was to spark ideas how to solve a physical problem such as race dynamics by leveraging the building blocks of #AI.
Thanks, Jason Cruz and team for organizing this event!
Melle de Wit, Marius Lucinskas, Paul Johnson, Valeria Loreti
This is awesome!