Turbo Lag Returned: Honda's Second ADUO Token, Singapore's Throttle and Aston Martin's Miscalculated Expectation

Honda spent its second ADUO token in Madrid on a new turbocharger. The component had one job: remove turbo lag, the delay between the driver pressing the throttle and the turbo delivering. A few weeks later at Marina Bay, Fernando Alonso had his foot flat and the car was answering late. After the race he pointed directly at power unit driveability. Honda sent engineers and power units back to Sakura. The deep checks found no hardware fault. What they found was the same old enemy — turbo lag. The question is not why the lag returned. The question is how a component, replaced at the cost of a token, loses its effect without breaking.

ADUO gets read as an ‘upgrade allowance’. It is really a rationing system. A token opens one defined hardware change window, and once spent, that window does not reopen. For a manufacturer playing catch-up it is the most expensive tool of the season, because after homologation it is the only route to touching hardware. Honda put its second token into the turbocharger. The logic was clean: the power unit spec introduced at Zandvoort already had weak driveability, and turbo lag was a leading cause. The token went where the pain was.

The pain came back at Marina Bay. Speaking to Motorsport.com, Honda Racing Corporation’s trackside general manager and chief engineer Shintaro Orihara admitted it plainly: even with the new turbocharger, managing lag in Singapore was hard because of the unique track layout and the energy-rich situation, and that gave the driver unstable acceleration. That sentence points away from the component and towards the control loop around it.

Understanding the energy-rich circuit matters, because that is where the whole thing jams. The MGU-K does two jobs — deploy, meaning deliver power, and harvest, meaning recover energy under braking and off-throttle. The traditional weapon against turbo lag is exactly that: harvest more in the off-throttle sections, keep the engine at a high operation point, keep the turbo spinning, and shorten the wait next time the throttle goes down. Orihara explained that this works only to a point. On energy-rich circuits it becomes hard to gather enough harvesting value, and when harvesting falls short, the engine’s operation mode cannot hold the lag. Two systems that normally cover for each other — the turbo and the MGU-K — fail in the same race.

The Sakura finding is as important as it is uncomfortable. No specific hardware fault. Nothing was broken; the way the power unit was being operated through settings and software simply could not meet the driver’s demand in those conditions. Orihara said lessons have been learned and countermeasures prepared, but those countermeasures live in control and settings, not in hardware. Aston Martin and Honda do not have an engine problem. They have an engine-map problem.

This is where the case meets track geometry. Turbo lag does not steal time on the straight; it steals time on corner exit. On a street circuit, where traction is limited and the walls are close, the driver cannot apply throttle linearly — he waits for the power, then builds the dose. The cost is bigger than the time lost: confidence in the braking point goes, the speed carried out of the corner changes, and preparation for the next corner suffers. The lap-time loss does not sit in one sector; it smears across the lap. Which is why telemetry shows not a slow car but an uneven one.

The second-order effect is underrated: race strategy. When a driver cannot push reliably on an out-lap, the undercut arithmetic weakens. I built the undercut index because the race was hiding in the pit window, and here the direction reverses — driveability uncertainty narrows the pit window. The team stretches the stint, declines the undercut, and becomes strategically defensive. That is what happened in Singapore. The pit wall spoke first; the points table only confirmed it later.

Mike Krack, Aston Martin’s chief trackside officer, said the thing that usually gets buried in a technical report as a footnote when it is actually the main text. Simulation, he admitted, cannot build a clear picture of driveability — how much lag or the MGU-K cutting out actually costs the driver goes beyond simple sensitivity modelling. One caveat here: the ‘driveability window’ I refer to is my own construct, built on six or seven energy-sensitive runs, and it should not be read as official timing data. Its signal is still clear: what the model cannot measure has to sit in a separate row of the risk register, and in Singapore that row was empty.

The contrarian side flips the story. The event was filed as a power unit failure, but after Sakura what emerged is a calibration-envelope problem — and that envelope was known since Zandvoort. Madrid’s token solved one variable called spool time; the software map solves the harvesting-deficit side. A new turbocharger is not a cure for the second problem. The team bought hardware where it needed a control loop. The replay shows the error; the structure shows why it was available.

Krack conceded the more uncomfortable cause himself, calling it ‘emotional dynamics’: Sepang over-delivered, expectations for Singapore rose beyond what was realistic, and once the weekend failed to deliver, everything got harder. He also accepted the team had not mapped a set of risks beforehand. That is a strategy-desk failure, not an engineering one. An energy-rich risk line should have been in the register before Marina Bay, with mitigations — a flatter throttle map, a track-specific harvesting mode — tested on Friday.

Austin is the next page, and the thing to watch there is not the scoreboard but the throttle trace. COTA’s energy profile differs from Marina Bay’s, so the question is not whether lag appears. The question is whether the software countermeasure makes the throttle linear or stepped. If the trace out of the slow corners in sector one looks like a staircase, Singapore’s problem was not solved, only postponed. And if Aston Martin uses the word ‘unexpected’ again in Austin, the risk register is still empty.