If you have ever opened a Smart Roadster MEG 1.1 tune and wondered how the ECU translates a torque demand into something the engine can actually act on, the answer lives inside a map called KFMIRL. This is the load conversion table at the heart of the Bosch torque-based control strategy, and understanding it is essential for anyone serious about kfmirl load map ecu tuning. Get it wrong and your other maps will fight each other, producing flat spots, unexpected boost surges or conservative fuelling that leaves power on the table. This article explains exactly what KFMIRL does, how it interacts with the rest of the torque structure, and what you need to consider when modifying it on the 698cc three-cylinder.
What Is KFMIRL and Where Does It Sit in the Torque Structure?
The Bosch MEG 1.1 uses a torque-based control architecture. Rather than commanding ignition timing, fuelling and boost directly, the ECU first calculates a torque request — a number in Newton-metres — and then converts that request into actionable signals for each actuator. KFMIRL is the map that performs one of the most critical steps in that chain: it converts the indicated torque request into a relative engine load value, typically expressed as a percentage of theoretical maximum cylinder filling.
This load value then feeds downstream maps governing injection duration, ignition advance and, indirectly, boost control. Think of KFMIRL as the translator sitting between the high-level torque demand and the low-level hardware commands. Without an accurate translation here, every other map in the ECU is working from a corrupted input. The axes of KFMIRL are typically engine speed (RPM) on one axis and requested torque on the other, with the output being the relative load figure the rest of the strategy will use.
For broader context on how all these maps coexist inside the Bosch firmware, the complete breakdown of the MEG 1.1 architecture is the best starting point before diving into individual tables.
KFMIRL vs KFMIOP: Understanding the Relationship
KFMIRL does not operate in isolation. It works in close partnership with KFMIOP, the optimal torque map. KFMIOP defines the maximum torque the ECU considers achievable at a given load and RPM combination — essentially the ceiling of the torque structure under normal operating conditions. KFMIRL then takes whatever torque value has been requested (which may be at, below or approaching that ceiling) and converts it into the load signal that downstream maps will use to calculate fuelling and spark.
The distinction matters enormously when tuning. If you raise boost and fuelling maps to support more power but leave KFMIRL unmodified, the ECU’s load calculation will no longer accurately reflect what is actually happening inside the cylinder. The result is a mismatch: the engine is producing more torque than the ECU believes it is, so ignition timing and injection may be calibrated for a lower-stress condition than actually exists. Conversely, an incorrectly scaled KFMIRL can cause the ECU to over-estimate load, triggering torque limiters or protective fuel enrichment unnecessarily.
Our dedicated article on how KFMIOP sets the optimal torque ceiling explains the upper boundary that KFMIRL must remain consistent with — read both together for the full picture.
How the 698cc Engine’s Characteristics Shape KFMIRL Values
The Smart Roadster’s 698cc three-cylinder is a physically small engine with a very high specific output relative to its displacement, particularly in Brabus trim. Its small swept volume means the load percentages encoded in KFMIRL can climb steeply even at moderate torque figures. A 698cc cylinder filling to theoretical maximum represents far less absolute air mass than a 1.6-litre unit would, so the ECU must scale its load calculations accordingly.
The Garrett 1238S turbocharger adds another layer of complexity. Because it is a fixed-geometry unit with a pneumatic wastegate, boost delivery is not as precisely controllable as a variable-geometry or electronic wastegate setup. This means that KFMIRL values at higher RPM, where the turbo is operating deep into its efficiency island, need particular care. The load signal must accurately reflect the denser charge entering the cylinder so that ignition timing does not advance into knock territory and fuelling remains stoichiometrically correct.
If you are new to the engine itself, understanding its fundamental architecture — displacement, firing order, cooling paths and the mid-rear mounting layout — is covered in depth in our guide to the 698cc three-cylinder’s core design, which provides essential context for any mapping work.
Practical Tuning: Modifying KFMIRL Safely
When to Modify KFMIRL
KFMIRL should be revisited any time you make a meaningful change to the engine’s torque output. This includes turbo upgrades, significant boost increases, camshaft changes or fuelling strategy revisions. If you are simply optimising ignition timing on a standard-boost car, KFMIRL may need only minor refinement. But if you are pushing beyond the 66kW SB2 calibration towards the kind of outputs our PRO or EVOLUTION maps target, the load conversion table will need rescaling to remain accurate.
Axis Scaling and Interpolation
The MEG 1.1 uses bilinear interpolation between map cells, so the resolution of your KFMIRL axis breakpoints is important. Too coarse a grid and the ECU will interpolate across a wide torque range with a single slope, producing load values that are accurate at the axis points but diverge meaningfully between them. When rescaling, maintain finer breakpoint spacing in the RPM ranges where the torque curve is steepest — typically 2,500 to 4,500 RPM on the boosted 698cc unit.
Checksum Correction After Edits
Any modification to KFMIRL, like all MEG 1.1 map edits, requires correct checksum recalculation before the firmware can be flashed successfully. Skipping this step results in an ECU that rejects the file entirely or, on some firmware revisions, accepts it but defaults to a limp-home strategy. The technical detail behind why this matters is explained in our article on checksum correction for the MEG 1.1.
KFMIRL and the Downstream Impact on the Softouch Gearbox
One aspect of KFMIRL that Smart Roadster tuners frequently overlook is its downstream effect on the Softouch automated manual gearbox. The transmission control unit receives a torque signal derived from the ECU’s internal calculations — a signal that is itself shaped by the load value KFMIRL produces. If your KFMIRL has been recalibrated to report higher load values to support a more powerful tune, the Softouch TCU may interpret this as a signal to alter shift behaviour, clutch engagement timing or torque reduction requests during gear changes.
On heavily modified cars, an incorrectly scaled KFMIRL can cause the gearbox to become more hesitant during upshifts because the TCU believes the engine is working harder than it needs to. Equally, it can suppress the ECU’s torque during shifts by more than necessary, negating some of the performance gain you have worked to achieve. Understanding how the Softouch actuator responds to increased torque signals is therefore a useful companion read when calibrating KFMIRL on a modified car.
KFMIRL Load Map ECU Tuning: Common Mistakes and How to Avoid Them
The most frequent error when editing KFMIRL is treating it as a linear scaler — simply multiplying all values by a constant factor proportional to the power increase. In practice, the relationship between torque request and relative load is not perfectly linear across the full RPM and torque range of the 698cc engine. At low RPM and low load, the map values tend to be compressed relative to what a linear extrapolation would predict, because throttle response and manifold dynamics are more influential than turbo output. At high RPM and high load, turbo efficiency and charge temperature become dominant, shifting the relationship again.
A second common mistake is editing KFMIRL without simultaneously verifying the consistency of related maps including the injection duration tables and the ignition advance maps. KFMIRL’s output feeds these directly; an inconsistency here can produce over-fuelling at light throttle or dangerously lean conditions at full load, neither of which will be obvious from a static check of the map values in isolation. Always validate changes on a rolling road or with wideband lambda logging before committing to a final calibration.
For tuners interested in how we structure our own calibrations across the BASIC 90 HP through to the EVOLUTION 125 HP maps, our remapping packages page explains the performance targets and the methodology behind each level of tune.
Summary
KFMIRL is not a glamorous map, but it is a foundational one. Accurate kfmirl load map ecu tuning is the difference between a calibration where all the other tables pull in the same direction and one where subtle conflicts between torque request, load signal and actuator output quietly erode performance and reliability. On the Smart Roadster’s 698cc three-cylinder — an engine already operating at high specific load in standard form — precision here is non-negotiable. Understand what KFMIRL is converting, verify it is consistent with KFMIOP and your fuelling strategy, recalculate your checksums, and your overall tune will be built on solid ground.
