The qualifying session at the 2026 Belgian Grand Prix at Spa-Francorchamps highlighted a new challenge for some Formula 1® drivers: the complex behaviour of their power units’ machine learning algorithms. These adaptive systems, introduced under the 2026 regulations, have shown to affect driver performance in ways beyond their control, particularly in energy deployment and power delivery during critical moments on track.
McLaren team principal Andrea Stella pointed out that Oscar Piastri’s qualifying deficit to teammate Lando Norris was largely due to the power unit’s algorithmic behaviour rather than any lack of driving skill. Piastri lost around two tenths of a second on the long straight between Stavelot and the Bus Stop Chicane, a phase where the power unit derates as electric energy depletes. Similar issues were observed at Mercedes, where George Russell faced difficulties compared to Kimi Antonelli, attributed to differences in how their power units managed energy deployment.
The 2026 Formula 1® power units employ machine learning algorithms that adjust energy deployment in real time, making performance less predictable and highly sensitive to small variations. Piastri’s limited running during the weekend—due to a hydraulic leak and a qualifying mistake—likely disrupted the power unit’s ability to learn and optimise, compounding his performance challenges. Drivers now need to focus extensively on optimising energy harvesting and braking techniques, which directly influence the power unit’s live calculations affecting braking points and cornering speeds.
Spa-Francorchamps’ high-speed layout, with fewer natural braking zones, intensifies these challenges under the 2026 regulations. The limited opportunities for energy harvesting make it harder for drivers to maintain consistent power unit performance across laps. This complexity adds a new layer to qualifying strategy and execution, where managing the power unit’s adaptive behaviour becomes as crucial as traditional driving skills.
At Mercedes, George Russell’s struggles relative to Kimi Antonelli further illustrate how power unit operation differences can influence lap times independently of driver input. These issues underline the growing importance of software and energy management in the current Formula 1® era, where the car’s systems play an increasingly significant role in on-track performance.
Andrea Stella’s assessment emphasises that the performance gaps seen during qualifying are not simply down to driver error but are linked to the evolving nature of power unit technology. The adaptive algorithms require consistent data and track time to optimise effectively, and any disruption can lead to unpredictable results. This dynamic is reshaping how teams and drivers approach qualifying and race preparation.
The situation at Spa-Francorchamps demonstrates the technical and operational challenges introduced by the 2026 power unit regulations. It highlights the intricate balance between driver skill and machine learning-driven power unit management, particularly on circuits where energy harvesting opportunities are limited. Understanding and adapting to these systems will be key for drivers aiming to extract maximum performance in qualifying sessions under the new rules.
