By the time an error is caught, bench time is already spent
Fixed test plans lock you in before you understand the system
Fixed test plans lock you in before you understand the system
Fixed test plans lock you in before you understand the system
By the time an error is caught,
bench time is already spent.
By the time an error is caught,
bench time is already spent.
Fixed test plans lock you in before you understand the system
By the time an error is caught,
bench time is already spent.
Secondmind for Calibration replaces the fixed, upfront test plan with one automated DoE loop – selecting the next test, updating the model, and optimizing toward the target as each result comes in. Engineers reach an explainable, validated result in far fewer tests – then refine and finalize the map directly within Secondmind, without returning to the bench.
The challenge
A fixed test plan can't tell where testing will actually matter — so cost and rework pile up in three predictable ways.
Exhaustive testing, regardless of value
Without the ability to adapt to what's already been learned, testing continues across the full space - even where the outcome is already clear.
Bench time lost that can't be recovered
Test cells cost $100s an hour to run. Every hour spent on a test that doesn't add value is money the program can't get back.
Costly rework, at the worst possible time
Poorly targeted measurements can produce a model that doesn't hold up when tested - forcing engineers back to the bench when the cost of rework is highest.
Exhaustive testing, regardless of value
Without the ability to adapt to what's already been learned, testing continues across the full space — even where the outcome is already clear.
Bench time lost that can't be recovered
Test cells cost $100s an hour to run. Every hour spent on a test that doesn't add value is money the program can't get back.
Costly rework, at the worst possible time
Poorly targeted measurements can produce a model that doesn't hold up when tested — forcing engineers back to the bench when the cost of rework is highest.
The Secondmind solution
Secondmind for Calibration directs every test toward the result, so engineers reach a validated model in far fewer tests, with far less rework.
Every test earns its place
Secondmind identifies the most informative measurement to take at each step - no upfront planning required
High-precision surrogate models
True system behaviour is separated from measurement noise, with quantified confidence at every point.
Better calibration decisions, consistently
Engineers can understand, explore, and act on the result directly, regardless of who is running the program.
Every test earns
its place
Secondmind identifies the most informative measurement to take at each step - no upfront planning required.
High-precision surrogate models
True system behaviour is separated from measurement noise, with quantified confidence at every point.
Better calibration decisions, consistently
Engineers can understand, explore, and act on the result directly, regardless of who is running the program.
Key features
Key features
Fewer tests, a map you can trust - here are some of the features that make it possible.
Fewer tests, a map you can trust -
here are some of the features that
make it possible.
Explore
Explore
Key features
Secondmind gives engineers different features to map the design space, compare candidates,
and commit with confidence.
Optimal Map
Optimal Map
Know the performance ceiling so any trade-off is a deliberate choice, backed by data.
Know the performance ceiling so any trade-off is a deliberate choice, backed by data.
Computes the optimal calibration map across the full operating range, balancing every performance target simultaneously.
Establishes a data-driven ceiling to measure every smoothing or trade-off decision against.
Built from defined constraints — not a generic result that needs reworking later.
Convergence Indicator
Know when to stop testing.
Real-time signal of model quality and its rate of improvement, updated after every test.
Flags the point where the model is accurate enough to trust - avoiding both premature stopping and over-testing.
Gives engineers a clear data-backed signal for when a result is ready to act on.
Convergence Indicator
Know when to stop testing.
Real-time signal of model quality and its rate of improvement, updated after every test.
Flags the point where the model is accurate enough to trust — avoiding both premature stopping and over-testing.
Gives engineers a clear data-backed signal for when a result is ready to act on.
Map Smoothing
Meet real-world driveability requirements.
Applies automatic or manual smoothing to balance peak performance against smoothness.
Makes the performance/smoothness trade-off explicit, rather than leaving it for downstream teams to resolve.
Produces a map ready for ECU implementation.
Map Smoothing
Meet real-world driveability requirements.
Applies automatic or manual smoothing to balance peak performance against smoothness.e.
Makes the performance/smoothness trade-off explicit, rather than leaving it for downstream teams to resolve.
Produces a map ready for ECU implementation.
Proven results
Delivers faster time-to-market, lower development costs, and better-performing results.
Faster time -to-market
59.7%
reduction in engineering hours on a production ICE calibration program at Mazda.
Lower development costs
65%
reduction in test cell occupancy time - from 4 weeks to 1.5 weeks — in a benchmark against an incumbent calibration tool at a Tier 1 supplier.
Better quality and performance
2.3x
more of the calibration space explored in that same program, without adding a single day to the schedule.
Proven results
Delivers faster time-to-market, lower development costs, and better-performing results.
Faster time -to-market
59.7%
reduction in engineering hours on a production ICE calibration program at Mazda.
Lower development costs
65%
reduction in test cell occupancy time - from 4 weeks to 1.5 weeks — in a benchmark against an incumbent calibration tool at a Tier 1 supplier.
Better quality and performance
2.3x
more of the calibration space explored in that same program, without adding a single day to the schedule.
The Secondmind advantage
What makes this possible - working in one system.
No bench time wasted in the wrong place
Secondmind Active Learning runs as an automated DoE loop - acquiring data, analyzing it, updating the model, and directing the next test to where uncertainty is highest - rather than following a fixed plan.
Signal, not noise.
Measurement noise is automatically separated from true system behaviour, so the model reflects what the system is actually doing - not measurement error.
A result any engineer can verify
The complexity is handled, so any engineer can understand a result, explore it, and act on it directly.
The Secondmind advantage
What makes this possible - working in one system.
No bench time wasted in the wrong place
Secondmind Active Learning runs as an automated DoE loop - acquiring data, analyzing it, updating the model, and directing the next test to where uncertainty is highest - rather than following a fixed plan.
Signal, not noise.
Measurement noise is automatically separated from true system behaviour, so the model reflects what the system is actually doing - not measurement error.
A result any engineer can verify
The complexity is handled, so any engineer can understand a result, explore it, and act on it directly.
Proven results
Delivers faster time-to-market, lower development costs, and better-performing results.
Faster time -to-market
59.7%
reduction in engineering hours on a production ICE calibration program at Mazda.
Lower development costs
65%
reduction in test cell occupancy time — from 4 weeks to 1.5 weeks — in a benchmark against an incumbent calibration tool at a Tier 1 supplier.
Better quality and performance
2.3x
more of the calibration space explored in that same program, without adding a single day to the schedule.
Proven results
Delivers faster time-to-market, lower development costs, and better-performing results.
Faster time -to-market
59.7%
reduction in engineering hours on a production ICE calibration program at Mazda.
Lower development costs
65%
reduction in test cell occupancy time — from 4 weeks to 1.5 weeks — in a benchmark against an incumbent calibration tool at a Tier 1 supplier.
Better quality and performance
2.3x
more of the calibration space explored in that same program, without adding a single day to the schedule.
Proven results
Delivers faster time-to-market, lower development costs, and better-performing results.
Faster time -to-market
59.7%
reduction in engineering hours on a production ICE calibration program at Mazda.
Lower development costs
65%
reduction in test cell occupancy time — from 4 weeks to 1.5 weeks — in a benchmark against an incumbent calibration tool at a Tier 1 supplier.
Better quality and performance
2.3x
more of the calibration space explored in that same program, without adding a single day to the schedule.
The Secondmind advantage
What makes this possible - working in one system.
No bench time wasted in the wrong place
Secondmind Active Learning runs an automated DoE loop — acquiring data, analyzing it, updating the model, and directing the next test to where uncertainty is highest — rather than following a fixed plan.
Signal,
not noise.
Measurement noise is automatically separated from true system behaviour, so the model reflects what the system is actually doing — not measurement error.
A result any engineer can verify
The complexity is handled, so any engineer can understand a result, explore it, and act on it directly.
Enterprise-ready
ISO 27001 certified. Cloud-native, browser-based, no new infrastructure.
Integrates with existing CAE workflows, including GT-Suite, MATLAB/Simulink, and JSOL JMAG.
Enterprise-ready
ISO 27001 certified. Cloud-native, browser-based, no new infrastructure.
Integrates with existing CAE workflows, including GT-Suite, MATLAB/Simulink, and JSOL JMAG.
Example use cases
Complex products, dozens of interacting variables, measurable impact.
E-motor calibration
Cut calibration time by 85% and test bench usage to a fifth of previous methods, with 80% fewer motor measurements.
ICE calibration
Mazda achieved a 59.7% reduction in engineering-hours to create high-precison calibration maps.
Diesel calibration
Reached minimum fuel consumption within Euro 6 limits in under 400 test points, vs. 2,500+ conventionally required.
E-powertrain calibration
Removed lengthy soaking periods needed to stabilize rotor temperature, cutting testbed occupancy by 80% - without compromising torque accuracy.
Example use cases
Complex products, dozens of interacting variables, measurable impact.
E-motor calibration
Cut calibration time by 85% and test bench usage to a fifth of previous methods, with 80% fewer motor measurements.
ICE calibration
Mazda achieved a 59.7% reduction in engineering-hours to create high-precison calibration maps.
Diesel calibration
Reached minimum fuel consumption within Euro 6 limits in under 400 test points, vs. 2,500+ conventionally required.
E-powertrain calibration
Removed lengthy soaking periods needed to stabilize rotor temperature, cutting testbed occupancy by 80% - without compromising torque accuracy.
Example use cases
Complex products, dozens of interacting variables, measurable impact.
E-motor calibration
Cut calibration time by 85% and test bench usage to a fifth of previous methods, with 80% fewer motor measurements.
ICE calibration
Mazda achieved a 59.7% reduction in engineering-hours to create high-precison calibration maps.
Diesel calibration
Reached minimum fuel consumption within Euro 6 limits in under 400 test points, vs. 2,500+ conventionally required.
E-powertrain calibration
Removed lengthy soaking periods needed to stabilize rotor temperature, cutting testbed occupancy by 80% - without compromising torque accuracy.
E-motor calibration
Cut calibration time by 85% and test bench usage to a fifth of previous methods, with 80% fewer motor measurements.
ICE calibration
Mazda achieved a 59.7% reduction in engineering-hours to create high-precison calibration maps.
Diesel calibration
Reached minimum fuel consumption within Euro 6 limits in under 400 test points, vs. 2,500+ conventionally required.
E-powertrain calibration
Removed lengthy soaking periods needed to stabilize rotor temperature, cutting testbed occupancy by 80% - without compromising torque accuracy.
Example use cases
Complex products, dozens of interacting variables, measurable impact.
E-motor calibration
Cut calibration time by 85% and test bench usage to a fifth of previous methods, with 80% fewer motor measurements.
ICE calibration
Mazda achieved a 59.7% reduction in engineering-hours to create high-precison calibration maps.
Diesel calibration
Reached minimum fuel consumption within Euro 6 limits in under 400 test points, vs. 2,500+ conventionally required.
E-powertrain calibration
Removed lengthy soaking periods needed to stabilize rotor temperature, cutting testbed occupancy by 80% - without compromising torque accuracy.
E-motor calibration
Cut calibration time by 85% and test bench usage to a fifth of previous methods, with 80% fewer motor measurements.
ICE calibration
Mazda achieved a 59.7% reduction in engineering-hours to create high-precison calibration maps.
Diesel calibration
Reached minimum fuel consumption within Euro 6 limits in under 400 test points, vs. 2,500+ conventionally required.
E-powertrain calibration
Removed lengthy soaking periods needed to stabilize rotor temperature, cutting testbed occupancy by 80% - without compromising torque accuracy.
Ready to calibrate with confidence?
See how Secondmind for Calibration could reduce bench time in your next program.
Ready to calibrate with confidence?
See how Secondmind for Calibration could reduce bench time in your next program.
Ready to calibrate with confidence?
See how Secondmind for Calibration could reduce bench time in your next program.
Insights
Stay ahead in a rapidly evolving industry. Access our latest resources, insights, and tools to equip you with the knowledge you need to stay competitive.
Insights
Stay ahead in a rapidly evolving industry. Access our latest resources, insights, and tools to equip you with the knowledge you need to stay competitive.
Insights
Stay ahead in a rapidly evolving industry. Access our latest resources, insights, and tools to equip you with the knowledge you need to stay competitive.
Insights
Stay ahead in a rapidly evolving industry. Access our latest resources, insights, and tools to equip you with the knowledge you need to stay competitive.
Product
Insights
Get in touch
© Secondmind 2025
Product
Insights
Get in touch
© Secondmind 2025
Product
Insights
Get in touch
© Secondmind 2025
Product
Insights
Get in touch
© Secondmind 2025
Product
Insights
Get in touch
© Secondmind 2025