Analytics and motorsport: a winning combination!

[#1]

How did you build a motorcycle that consistently achieves podium finishes in various competitions? It is not without significance, it turns out, is by using analytics to improve its performance. A team from Politechnika Warszawska built their own motorbike, then worked with SAS to gather and analyse data about its performance in training sessions and during races. They attributed their well-deserved third place at Poznań recently—from a starting point on the grid of ninth place—to using analytics to fine-tune various parameters. I spoke to Paweł Chodkiewicz and Jan Biniewicz about the work of the team, and how they used analytics to improve the performance of their bike.

Tell me more about the project. What was the goal?

Our main goal was to apply the theoretical background knowledge that we acquired during our studies to a real industrial project. We started our journey in motorcycle design and racing in 2017, when we registered our team for the academic MotoStudent competition. We completed our first bike in April 2019, and later took it to two race weekends at Slovakiaring in Slovakia (achieving podium finishes) and one in Great Britain. Our new motorcycle was unveiled in September 2023, and it has been adapted to the regulations of the professional PreMoto3 and Moto4 racing series.

How big is the team?

We are a small team with 11 members from various faculties of Politechnika Warszawska. The home department is the Faculty of Automotive and Construction Machinery Engineering. The team is divided into departments of geometry and chassis design, electronics, aerodynamics, engine, and media.

Tell me more about the bike. Let’s have some technical details!

We are using a stock (series), non-internally modified engine from a sport road motorcycle, the KTM RC250. The engine has a capacity of 250 cm³ and generates 31 HP and 24 Nm of torque. We are using a standalone engine control unit), the Ecumaster EMU Black, which means we can customise the injection and ignition map and control all the engine parameters. The engine itself is non-internally modified, but we have developed our own exhaust and ram-air intake system. We use a 100 high-octane fuel, and the frame is our own design. It is a lightweight aluminium alloy twin-spar type chassis. Frame pieces were machined from over 100 kg of solid blocks and later welded, resulting in a 6.5 kg chassis ready to use. We can change the caster angle, frame length, and swingarm axle position using various inserts. The tire size is the same as in the Moto3 class, 90/80R17 in the front and 115/75R17 in the rear. The dry weight of the bike is 102 kg, going up to 113 kg with a full tank of fuel. The aerodynamic package was developed using Computational Fluid Dynamics (CFD) in collaboration with the Academy of Fine Arts in Warsaw, and was manufactured using a carbon/epoxy composite.

How do you collect information?

We have a Raspberry Pi 4 mini-computer connected to the dashboard to provide data, and more than 30 sensors attached to the bike. We use linear potentiometers to gather data about suspension and steering positions, GPS for positioning and velocity measurement, magnetoresistance sensors for wheel rotational velocity, pyrometers (three per wheel) for tire temperature, and so on. We also have an accelerometer with a gyroscope, a brake pressure sensor, and all the data from the engine control unit, such as engine speed, throttle position sensor, and water and oil temperature

.

How do you analyse the data that you collect?

Since we started our partnership with SAS, we have not needed to use any other software to analyse our data. We use SAS Event Stream Processing for real-time data streaming and processing, SAS Studio for data preparation, and SAS Visual Analytics for interactive data exploration. Within the Visual Analytics platform, we built specific templates tailored to different areas of interest to use when working on the racetrack. These templates cover tasks such as motorcycle trim optimisation, rider control, grip management, and engine health. We can also compare laps within the same training session and across sessions. Our results at Poznań show that being able to tweak the bike’s parameters across sessions can make all the difference!

What are your future plans for the project?

We plan to collect as much data as possible with the current bike to prepare us for a new experience and challenge: developing an electric motorcycle for the next MotoStudent competition. We also have plans to participate in a selected round of the Moto4 class in the Trofei MES series in Italy in 2024.