9 hours ago
Hi everyone,
I have been discussing with ChatGPT an idea that started with a fairly simple question but has gradually turned into what I think could be an interesting SPA/OrBit reverse-engineering project:
Could a MY2024 T6 Recharge B4204T52 be given a genuine “Polestar-like” optimisation — not simply more power, but the complete powertrain behaviour that Polestar normally changes — even though Volvo/Polestar does not offer the optimisation for this engine?
My car is a MY2024 XC60 T6 Recharge (B4204T52, 350 hp system output).
I should give ChatGPT credit for most of the research below. I initially asked whether AI could theoretically create a Polestar-type optimisation for a Volvo for which Polestar does not offer one. ChatGPT then researched T6/T8 hardware, Volvo part numbers, ECM references, OrBit parameters and previous discussions on this forum. We have subsequently also gone through technical information and owner discussions concerning the MY22.5+ Extended Range powertrain.
What follows is our current working hypothesis. Some part-number matches come from Volvo documentation, parts catalogues and donor vehicles rather than engineering documentation, so I am deliberately treating them as evidence rather than proof.
1. The interesting comparison is specifically the MY22.5+ Extended Range architecture
The first important conclusion was that comparing “T6 vs T8” generically is misleading.
The relevant family is the newer Extended Range / VEA2 architecture.
For MY2024 we have:
B4204T52 — T6 Recharge
The MY22.5 B4204T57 is also an interesting reference. It appears to have been an early Extended Range implementation, with essentially T8-level output, ERAD 7 and the C-ISG, before the later B4204T56 became the relevant engine code in many markets.
This gives us another useful reference point rather than treating T52/T56 as an isolated pair.
2. B4204T52 vs B4204T56 hardware
ChatGPT cross-checked the available documentation and part numbers and found a surprisingly large amount of apparent common hardware.
Among the matches/equivalences found so far:
There could still be differences in pistons, rods, internal cooling, valvetrain or other components.
However, there seems to be enough common hardware to make the 253/350 → 310/400 difference very interesting from a calibration perspective.
3. Same ECM hardware
Another potentially important finding was examples of MY2024 B4204T52 and B4204T56 cars apparently using the same Volvo/Denso ECM hardware:
32383818 / Volvo 36010797
If that is correct, we potentially have:
very similar engine architecture + same ECM hardware + different rated engine output.
What we do not yet know is whether the software is:
A) identical, with the engine/power variant selected through configuration;
B) the same executable with different calibration blocks; or
C) genuinely different ECM software builds.
That distinction seems fundamental.
4. OrBit makes the question particularly interesting
OrBit exposes 005 ENGINE with specific engine power/torque variants, including values corresponding to MP and HP versions, and also exposes 368 ENGINE FAMILY TYPE.
I also found the previous discussion here concerning a T6 Recharge B4204T52 → T8 B4204T56 conversion.
x119's response — that comparing the actual ECM software would be interesting and that, if the software were identical, configuration might potentially select another mapping — is very close to the conclusion ChatGPT and I independently reached.
But I want to stress that simply changing 005 ENGINE is not what I am proposing.
If the ECM does not contain the corresponding calibration, changing a configuration descriptor obviously cannot magically create it.
The interesting question is what the relationship actually is between configuration and calibration on these cars.
5. There may be another important difference: hybrid torque/power blending
This is the part of the project that now interests me even more than the ICE power difference.
Volvo rates the MY24 T6 as:
253 hp ICE + 145 hp ERAD → 350 hp system
while the T8 is:
310 hp ICE + 145 hp ERAD → 455 hp system
I realise that individual motor peak outputs cannot automatically be added. They may occur at different rotational speeds/vehicle speeds and there may be battery, inverter and drivetrain constraints.
So I am not assuming that the T6 has a simple electronic “350 hp limiter”.
However, the difference raises an interesting question:
Is the T6/T8 distinction partly a difference in hybrid torque blending and allowed simultaneous ICE/ERAD contribution, rather than only a 253 → 310 hp ECM calibration difference?
If so, converting or optimising the ICE alone would tell only part of the story.
6. And there is also the C-ISG
The Extended Range technical information adds another component that I initially underestimated.
The crankshaft-integrated starter/generator (C-ISG) is substantial and Volvo documentation describes it as working together with the turbo in low-speed operation.
This means that actual wheel-torque delivery is potentially a coordinated result of:
ICE + C-ISG + ERAD + battery/inverter limits + transmission + hybrid torque management.
There are also owner reports suggesting that C-ISG assistance may be particularly noticeable transiently at low speeds / during launch.
I therefore wonder whether a proper comparison of T52/T56 should include not only ECM and TCM but also whatever module/configuration determines C-ISG and ERAD torque requests.
7. Why simply enabling “Polestar” is not the objective
A genuine Polestar Engineered Optimisation is more than increased engine output.
Depending on application, Polestar changes things such as:
In other words, Polestar itself demonstrates that an optimisation can materially improve the car without increasing the headline horsepower figure.
That is much closer to what I am interested in.
8. The experiment I would really like to see
Rather than inventing a tune, I think a more rigorous approach would be to obtain several OEM reference states:
A — B4204T52 T6 OEM
B — B4204T56 T8 OEM
C — B4204T57 Extended Range, where useful as an additional reference
D — B4204T56/T57 T8 with genuine Polestar optimisation
Then perform two principal comparisons:
Diff 1: T52 OEM → T56 OEM
This should tell us how Volvo differentiates the contemporary T6 and T8.
Not just in ECM power calibration, but potentially also in hybrid torque coordination.
Diff 2: T8 OEM → T8 Polestar
This should tell us how Volvo/Polestar optimises the behaviour of an otherwise identical powertrain.
The second comparison may actually be the more valuable one.
The question would then become:
Can the transformation identified in Diff 2 be transferred, where technically applicable, to the B4204T52 architecture?
In other words, not:
“Make a T6 into a T8.”
but:
“Apply the engineering philosophy used to transform a T8 into a T8 Polestar to the T6.”
9. Where AI could be useful
This is where my discussion with ChatGPT originally started.
If sufficiently complete T52/T56/Polestar software and configuration data could be obtained, AI-assisted analysis could potentially help identify:
AI would instead be used as a reverse-engineering assistant to answer:
What did Volvo change between T52 and T56?
and:
What did Polestar change between OEM T8 and Polestar T8?
Those known OEM transformations would then define the boundaries of any hypothetical T6 implementation.
10. The target is behaviour, not a horsepower number
Earlier in this investigation we discussed possible ICE outputs around 290–300 hp / ~400 Nm.
I now think defining the project that way misses the point.
I would rather have a T6 with a modest increase in power that behaves exactly as a hypothetical factory Polestar T6 might have behaved than a 450–500 hp aftermarket Stage 1.
The desired result would be something like:
not a tuned T6, but an attempt to reconstruct the T6 Polestar that Volvo never made.
11. Questions for x119 / the OrBit community
I would therefore be very interested in your thoughts on the following:
At this stage I am much more interested in characterising how Volvo differentiates T6/T8 and how Polestar transforms the complete powertrain behaviour than in blindly flashing anything into the car.
The research so far suggests that this could be an interesting SPA/OrBit reverse-engineering exercise rather than simply another chiptuning project.
I would be very interested to hear whether anyone here has already investigated any of these layers further — particularly the relationship between 005 ENGINE, ECM calibration and hybrid torque management on the Extended Range cars.
P.S. I realise this is a very long post, but I wanted to provide enough context for anyone interested to be able to take the same information and start a conversation with whichever AI agent they use, and perhaps explore the problem independently.
I haven't seen much discussion here about using AI for this kind of reverse-engineering work — although I may simply have missed it. If it hasn't really been tried yet, I think it opens up some genuinely interesting possibilities, particularly when the task is not to invent a tune from scratch, but to compare known OEM implementations, identify what Volvo/Polestar actually changed, and reason from those differences.
I have been discussing with ChatGPT an idea that started with a fairly simple question but has gradually turned into what I think could be an interesting SPA/OrBit reverse-engineering project:
Could a MY2024 T6 Recharge B4204T52 be given a genuine “Polestar-like” optimisation — not simply more power, but the complete powertrain behaviour that Polestar normally changes — even though Volvo/Polestar does not offer the optimisation for this engine?
My car is a MY2024 XC60 T6 Recharge (B4204T52, 350 hp system output).
I should give ChatGPT credit for most of the research below. I initially asked whether AI could theoretically create a Polestar-type optimisation for a Volvo for which Polestar does not offer one. ChatGPT then researched T6/T8 hardware, Volvo part numbers, ECM references, OrBit parameters and previous discussions on this forum. We have subsequently also gone through technical information and owner discussions concerning the MY22.5+ Extended Range powertrain.
What follows is our current working hypothesis. Some part-number matches come from Volvo documentation, parts catalogues and donor vehicles rather than engineering documentation, so I am deliberately treating them as evidence rather than proof.
1. The interesting comparison is specifically the MY22.5+ Extended Range architecture
The first important conclusion was that comparing “T6 vs T8” generically is misleading.
The relevant family is the newer Extended Range / VEA2 architecture.
For MY2024 we have:
B4204T52 — T6 Recharge
- ICE: 253 hp / 350 Nm
- ERAD: 107 kW (145 PS) / 309 Nm
- system output: 350 hp
- ICE: 310 hp / 400 Nm
- ERAD: same 107 kW / 309 Nm
- system output: 455 hp
The MY22.5 B4204T57 is also an interesting reference. It appears to have been an early Extended Range implementation, with essentially T8-level output, ERAD 7 and the C-ISG, before the later B4204T56 became the relevant engine code in many markets.
This gives us another useful reference point rather than treating T52/T56 as an isolated pair.
2. B4204T52 vs B4204T56 hardware
ChatGPT cross-checked the available documentation and part numbers and found a surprisingly large amount of apparent common hardware.
Among the matches/equivalences found so far:
- 1,969 cc
- 82 × 93.2 mm bore/stroke
- 10.3:1 compression ratio
- same 107 kW / 309 Nm ERAD
- same 18.8 kWh battery architecture
- same TG-81SD transmission family
- evidence of the same BorgWarner turbo
- evidence of the same cylinder-head reference
- evidence of common direct-injection components/injectors
- evidence of the same high-pressure fuel pump
There could still be differences in pistons, rods, internal cooling, valvetrain or other components.
However, there seems to be enough common hardware to make the 253/350 → 310/400 difference very interesting from a calibration perspective.
3. Same ECM hardware
Another potentially important finding was examples of MY2024 B4204T52 and B4204T56 cars apparently using the same Volvo/Denso ECM hardware:
32383818 / Volvo 36010797
If that is correct, we potentially have:
very similar engine architecture + same ECM hardware + different rated engine output.
What we do not yet know is whether the software is:
A) identical, with the engine/power variant selected through configuration;
B) the same executable with different calibration blocks; or
C) genuinely different ECM software builds.
That distinction seems fundamental.
4. OrBit makes the question particularly interesting
OrBit exposes 005 ENGINE with specific engine power/torque variants, including values corresponding to MP and HP versions, and also exposes 368 ENGINE FAMILY TYPE.
I also found the previous discussion here concerning a T6 Recharge B4204T52 → T8 B4204T56 conversion.
x119's response — that comparing the actual ECM software would be interesting and that, if the software were identical, configuration might potentially select another mapping — is very close to the conclusion ChatGPT and I independently reached.
But I want to stress that simply changing 005 ENGINE is not what I am proposing.
If the ECM does not contain the corresponding calibration, changing a configuration descriptor obviously cannot magically create it.
The interesting question is what the relationship actually is between configuration and calibration on these cars.
5. There may be another important difference: hybrid torque/power blending
This is the part of the project that now interests me even more than the ICE power difference.
Volvo rates the MY24 T6 as:
253 hp ICE + 145 hp ERAD → 350 hp system
while the T8 is:
310 hp ICE + 145 hp ERAD → 455 hp system
I realise that individual motor peak outputs cannot automatically be added. They may occur at different rotational speeds/vehicle speeds and there may be battery, inverter and drivetrain constraints.
So I am not assuming that the T6 has a simple electronic “350 hp limiter”.
However, the difference raises an interesting question:
Is the T6/T8 distinction partly a difference in hybrid torque blending and allowed simultaneous ICE/ERAD contribution, rather than only a 253 → 310 hp ECM calibration difference?
If so, converting or optimising the ICE alone would tell only part of the story.
6. And there is also the C-ISG
The Extended Range technical information adds another component that I initially underestimated.
The crankshaft-integrated starter/generator (C-ISG) is substantial and Volvo documentation describes it as working together with the turbo in low-speed operation.
This means that actual wheel-torque delivery is potentially a coordinated result of:
ICE + C-ISG + ERAD + battery/inverter limits + transmission + hybrid torque management.
There are also owner reports suggesting that C-ISG assistance may be particularly noticeable transiently at low speeds / during launch.
I therefore wonder whether a proper comparison of T52/T56 should include not only ECM and TCM but also whatever module/configuration determines C-ISG and ERAD torque requests.
7. Why simply enabling “Polestar” is not the objective
A genuine Polestar Engineered Optimisation is more than increased engine output.
Depending on application, Polestar changes things such as:
- throttle response
- engine torque delivery
- shift speed
- shift points
- gear-hold behaviour after throttle lift
- drivetrain torque management
- AWD behaviour / torque distribution
In other words, Polestar itself demonstrates that an optimisation can materially improve the car without increasing the headline horsepower figure.
That is much closer to what I am interested in.
8. The experiment I would really like to see
Rather than inventing a tune, I think a more rigorous approach would be to obtain several OEM reference states:
A — B4204T52 T6 OEM
B — B4204T56 T8 OEM
C — B4204T57 Extended Range, where useful as an additional reference
D — B4204T56/T57 T8 with genuine Polestar optimisation
Then perform two principal comparisons:
Diff 1: T52 OEM → T56 OEM
This should tell us how Volvo differentiates the contemporary T6 and T8.
Not just in ECM power calibration, but potentially also in hybrid torque coordination.
Diff 2: T8 OEM → T8 Polestar
This should tell us how Volvo/Polestar optimises the behaviour of an otherwise identical powertrain.
The second comparison may actually be the more valuable one.
The question would then become:
Can the transformation identified in Diff 2 be transferred, where technically applicable, to the B4204T52 architecture?
In other words, not:
“Make a T6 into a T8.”
but:
“Apply the engineering philosophy used to transform a T8 into a T8 Polestar to the T6.”
9. Where AI could be useful
This is where my discussion with ChatGPT originally started.
If sufficiently complete T52/T56/Polestar software and configuration data could be obtained, AI-assisted analysis could potentially help identify:
- identical vs differing executable regions
- calibration blocks
- torque-model tables
- requested/allowed ICE torque
- load targets
- boost targets
- fuelling/lambda
- ignition
- thermal protection
- gear-dependent torque limits
- pedal → torque-request relationships
- shift-related torque intervention
- ICE/C-ISG/ERAD torque coordination
- differences in hybrid torque blending
AI would instead be used as a reverse-engineering assistant to answer:
What did Volvo change between T52 and T56?
and:
What did Polestar change between OEM T8 and Polestar T8?
Those known OEM transformations would then define the boundaries of any hypothetical T6 implementation.
10. The target is behaviour, not a horsepower number
Earlier in this investigation we discussed possible ICE outputs around 290–300 hp / ~400 Nm.
I now think defining the project that way misses the point.
I would rather have a T6 with a modest increase in power that behaves exactly as a hypothetical factory Polestar T6 might have behaved than a 450–500 hp aftermarket Stage 1.
The desired result would be something like:
- sharper but OEM-like throttle response
- stronger mid-range ICE delivery
- appropriate C-ISG contribution
- better ICE/ERAD coordination
- faster and more appropriate downshifts
- better gear holding
- Polestar-like shift strategy
- appropriate drivetrain/AWD behaviour
- all OEM thermal/component protections retained
not a tuned T6, but an attempt to reconstruct the T6 Polestar that Volvo never made.
11. Questions for x119 / the OrBit community
I would therefore be very interested in your thoughts on the following:
- Has anyone compared the actual ECM software/calibration IDs of late B4204T52 and B4204T56 cars?
- Does anyone know whether the 32383818 T52/T56 ECMs contain the same executable with different calibration blocks?
- Could 005 ENGINE / 368 ENGINE FAMILY TYPE select calibration behaviour already present in the ECM, or are these effectively configuration descriptors only on this generation?
- Has anyone captured a complete before/after software/configuration comparison from a genuine Extended Range T8 Polestar installation — particularly ECM, TCM and CEM configuration?
- On ERAD-based Recharge vehicles, which parts of Polestar behaviour appear to come from configuration and which definitely require flashed software?
- Has anyone compared the hybrid torque/power blending strategy between B4204T52 T6 and B4204T56 T8?
- Which module actually determines the maximum simultaneous ICE/C-ISG/ERAD torque request on these Extended Range cars?
- Is C-ISG contribution configurable/calibratable independently, and has anyone logged its requested/actual torque under full acceleration?
- Would the OrBit developers/community be interested in analysing a matched set of T52 OEM, T56 OEM and genuine T8 Polestar software/configuration data if it could be obtained?
At this stage I am much more interested in characterising how Volvo differentiates T6/T8 and how Polestar transforms the complete powertrain behaviour than in blindly flashing anything into the car.
The research so far suggests that this could be an interesting SPA/OrBit reverse-engineering exercise rather than simply another chiptuning project.
I would be very interested to hear whether anyone here has already investigated any of these layers further — particularly the relationship between 005 ENGINE, ECM calibration and hybrid torque management on the Extended Range cars.
P.S. I realise this is a very long post, but I wanted to provide enough context for anyone interested to be able to take the same information and start a conversation with whichever AI agent they use, and perhaps explore the problem independently.
I haven't seen much discussion here about using AI for this kind of reverse-engineering work — although I may simply have missed it. If it hasn't really been tried yet, I think it opens up some genuinely interesting possibilities, particularly when the task is not to invent a tune from scratch, but to compare known OEM implementations, identify what Volvo/Polestar actually changed, and reason from those differences.