Laptime Optimisation by Canopy Simulations
A Revolution in Motorsport Laptime Simulations
Configuring and solving optimal control (trajectory optimisation) problems using collocation.
Access Unparalleled Expertise
We are the only simulation supplier in Formula One, and the leading supplier to Formula E. We have over 35 years’ experience in the biggest Formula One teams. We also supply many other series including OEMs and regulatory bodies. We joined the Michelin Group of companies in April 2023.
Gain Lap Time, Faster
Use our scalable compute cloud to run large-scale multi-dimensional studies in minutes, while our cutting edge vehicle model will ensure your car is simulated with unprecedented accuracy.
Cut Your Costs
With Canopy Simulations you’ll get an in-depth understanding of your car’s performance in a fraction of the time, and at a fraction of the cost, of a typical in-house solution.
We built the state-of-the-art, so you don’t have to
At Canopy Simulations we observed that team after team will re-invent the same laptime simulation technology, with varying degrees of success. We realised that we could end the wasteful “not invented here” mentality, and get vehicle dynamicists back to focusing on what they do best; finding lap time.
We combined our decades of expertise to create the most advanced car model, simulations and data analysis tools yet seen in Formula One or any other motorsport division. Then we unleashed their true potential using the power of cloud computing.
The Canopy Platform has been built around these tools, enabling anyone to leapfrog the competition to have best-in-class simulation capability at a fraction of the cost of developing in-house.
Turning energy into lap time is one of the hardest problems in motorsport.
This video tells the inside story of how Nissan Formula E team took a deep dive into the regs and used a second motor spinning at 100,000rpm to gain ~𝟬.𝟴𝘀 𝗽𝗲𝗿 𝗹𝗮𝗽. Making it work wasn’t easy: multiple energy flows, complex control, and tight thermal limits. That’s where Canopy optimal control came in.
Light User
Our base subscription includes access to all of the platform’s core functionality.
Inclusive of 5000 Compute Credits per month, 120000 Storage Credits, and up to 6 available cloud nodes. We do not place a limit on the number of users per subscription.
Medium, Heavy & Super User
These packages include greater numbers of Compute-Credits and Storage-Credits, and available cloud nodes.
We can also create a custom package based on your team’s individual requirements.
Popular Add-ons
Optional-credit-rollover lets you to annualise your Compute Credits – great if your usage requirements will fluctuate. You can upgrade to 8-Nodes-Always-On if you’ll be running lots of small studies simultaneously and will require faster execution times.
We also offer use of the Canopy Vehicle Model in Driver-in-the-loop applications, supplied in a compiled form, compatible with your simulator framework.
Our Simulations
Dynamic Lap
The holy grail of laptime simulation. A fully dynamic simulation with no driver model, achieving the maximum performance of the car, including all of its dynamics, without any assumptions about how it is driven.
Dynamic Lap solves every state of the car model at every point on the track simultaneously to give a dynamically consistent state trajectory which traverses the lap in minimum time subject to any number of algebraic or integral constraints on its behaviour.
With support for multiple laps, powertrain thermal and tyre thermal, Dynamic Multi-Lap allow you to optimise entire stints.
This fully Dynamic laptime simulation has made Quasi-static simulation largely obsolete. Dynamic Lap is the easiest way to run a fully Dynamic simulation, with ability to vary conditions around the lap such at wind and elevation, impose energy/thermal constraints, and optimise numerous control inputs for instance torque vectoring, brake balance, active ride height.
Secondary Lap Sensitivities
The computation of local sensitivities around the lap help us to evaluate complex engineering trade-offs. Additional vector outputs are computed for lift and drag coefficients, ride heights, mass, power and grip.
When we’re deciding whether to fit a component or make a setup change, we typically gain in some places, but lose in others. Secondary Lap Sensitivities allow you to identify where these gains and losses originate.
Channel Inference
Working backwards from telemetry data, this simulation uses the vehicle model to estimate the value for channels which were not measured directly. It uses the quasi-static equilibrium of the car and an inverse tyre model to make these deductions.
Channel Inference doesn’t have to be run on track data; any laptime simulation output can be staged as telemetry. In this way it is possible to look at the effect of small changes in car setup on balance.
For example if two inference sims are run with different rear anti-roll bar rising rates on the same telemetry we could examine delta rLatBalF in the results to see how this might affect car balance at particular points around the circuit.
While this is a simplification, because we wouldn’t achieve exactly the same vCar-gLat-gLong around the circuit after the setup change, it can be easier to interpret than a laptime simulation, because the lap driven doesn’t change.
Racing Line Optimisation
Given a car and a track definition containing only the x-y and optionally z coordinates of the track edges, Dynamic Lap will automatically find the racing line which enables the given car to lap the track in the shortest time.
Alternatively a fixed racing line can be defined which might be useful for correlation to track data.
Straight Sim
Straight Sim is a straight-line simulation which finds the static equilibrium state (with no acceleration) of the car at a series of speeds from 0 to 100m/s.
Tyre Thermal Dynamic Lap
After correlating to track or rig data using Thermal Replay, the thermal dynamics of the tyre ribs, through thickness elements, wheels, and brakes can be computed.
Combined with the effect of temperature on grip and cornering stiffness, Dynamic Lap can find the optimal driving style to get the most out of the tyres. This allows us to make informed decisions regarding the thermal effect of setup changes.
Combined with strategy weights this tool can also be used to find tyre saving opportunities around the lap while maintaining a competitive advantage in places which are critical for overtaking.
This is a computationally demanding exercise with 250,000 unknowns, but thanks to the powerful cloud computing resources available, up to 200 laps can be computed in parallel in around 25 minutes.
This tyre thermal model can be run in real time in the driver-in-loop simulator so drivers experience the feedback effect of their driving on tyre temperature.
Thermal Replay
In order to speed up correlation to your measured powertrain or tyre temperatures this tool enables you to input telemetry data, run it through the thermal model, and examine your fit to track data.
In conjunction with explorations you can sweep through thousands of combinations of parameters in minutes to help you quickly dial in your correlation.
If you’re interested in powertrain temperatures and battery derating, then you can run these parameters through Dynamic Multi-Lap to perform a full stint energy and thermal optimisation. Alternatively if you are concerned about tyre temperatures, then this puts you in a good place to run Tyre Thermal Dynamic Lap.
Virtual N-Post Rig
The car is run over a defined road profile by feeding the road heights into the four corners (wheels) of the car in continuous time.
The results show the dynamic response of the car to the given road profile, enabling analysis of contact patch load variation and exploration of how changes in damping and compliance will affect ride characteristics.
Ride Sim
Ride Sim is used for assessing car performance over track bumps. Ride Sim will firstly compute the optimal racing line for a given car, before driving the car around this racing line over a fully featured track bump profile. Users can then assess the ride performance metrics of the car at the limits of performance, taking into account the inertial loads, the impact of aerodynamics on ride and all other effects on the vertical dynamics we would want to include.
Cornering Sim
The cornering version of Straight Sim. User can customise the gLat-gLong angle for which quasi-static equilibrium is achieved. The simulation will then return the maximum cornering performance for the customised condition (e.g. late entry, mid corner, early exit…) through the speed range.
Developing Capabilities
Canopy Simulations is continuously expanding the tools available on the platform. The capabilities below are just some that are in active development, designed to give motorsport teams and vehicle dynamics engineers better ways to build and optimise vehicle models.
Cosimulation
Cosimulation allows external "black box" components to run alongside the Canopy vehicle model during a Dynamic Lap simulation. A team can connect a proprietary "subsystem" to Canopy's optimal control solver without revealing the subsystem's internal logic or intellectual property. This could be an aero map, a tyre model, a powertrain controller, an energy management model, or any other component the team wants to keep in house.
Canopy Simulations already supports external model integration: either implementing a client's model within the Canopy codebase; or interfacing directly with a client produced model binary. The Cosimulation capability will extend this further by enabling direct coupling of third party components within the simulation loop.
Dynamic Inference
Dynamic Inference is a capability that derives unmeasured vehicle channels from logged telemetry data using a fully dynamic vehicle model. It extends Canopy's existing Channel Inference toolset, replacing quasi-static assumptions with Canopy's direct collocation solver to deliver greater accuracy and control over the results. The outcome is a more reliable reconstruction of vehicle behaviour from measured track data.
TameTire Dynamic Lap
Canopy Simulations offers a range of tyre modelling approaches on the platform. To add to these options, Canopy is developing the ability to run Michelin's TameTire tyre model directly inside a fully dynamic laptime optimisation. TameTire is widely regarded as the state of the art in racing tyre modelling, and coupling it with Canopy's cloud native optimal control solver will capture tyre behaviour with a level of physical detail that simpler tyre models cannot represent. This combination is not available in any other simulation environment.
Laptime Simulation for every racing series
Canopy supports vehicle architectures from GT cars and endurance prototypes to single-seaters, with standard ICE, electric and complex hybrid powertrains. It can model everything from conventional race circuits and banked ovals to more demanding layouts like the Nordschleife or Pikes Peak. Canopy’s models evolve alongside major regulatory changes, enabling teams to assess new vehicle architectures and operating constraints early in the development cycle.
Formula 1 2026
The 2026 regulations marked a major shift in the history of F1 power units, giving much more importance to the electric powertrain and energy management.
These changes have created a highly complex hybrid system with different operating modes: Boost, Overtake, part-throttle regen, super-clipping…
Canopy models these operating modes within a unified simulation environment, enabling engineers to analyze their interactions over a complete lap and evaluate the performance impact of different control strategies.
Formula E Gen 4
Formula E is a racing category where energy management is critical.
The limited amount of available energy requires teams to deploy it carefully.
Our Storage model is designed to replicate your actual battery and our solver helps you find the optimal way to use that energy.
With brakes on all 4 corners and electric motors on both axles, there are many possible ways to achieve deceleration while harvesting energy and cornering.
Our “Optimal BBW with max regen” control strategy is designed to let the solver choose the optimal distribution between front and rear, and between brakes and motors.
Gen 4 regulation also enables active differentials on both axles.
As with BBW, you can let the solver choose the diff characteristics thanks to our “Optimal Diff Torque” setting.
Hypercar – WEC LMH and IMSA GTP
In the WEC and IMSA endurance championships, the Balance of Performance (BoP) imposes specific constraints on each car.
The overall power delivered to the wheels is limited, and the energy that can be deployed over a stint is also capped (also known as the “Virtual Energy Tank” in WEC).
These regulatory constraints are covered in Canopy through our “Powertrain Limits” feature so that you can anticipate the impact of a BoP adjustment.
Endurance racing is not just about peak performance, success also depends on finding a car setup which remains competitive through multiple stints while managing tyre wear and degradation. This is especially important since, unlike most top-level racing series, tyre blankets and ovens are banned, meaning tyres must be brought up to temperature on track. To address this balance, Canopy offers the capability to simulate tyre thermals, degradation and multi-lap performance.
And more!
Our tool also covers most major racing series:
Lots of suspension topologies available
Ability to run fully asymmetric setups for Indycar and oval/banked circuits
Active aero and suspension components for advanced sports-car development programs
As a wise man once said, “If it’s got 4 wheels, we can simulate it”!
Check out: https://www.youtube.com/watch?v=7u9OtZV1vY8
Canopy provides a set of default car models to its customers to help them start running simulations and have guidelines on how to use features.
We have F1, F3, FE, Formula Student, Hypercar, Super GT and GT3 examples.
Contact our team to discuss your vehicle architecture, regulatory constraints and simulation objectives.
See it in action
FAQs
How secure is our team's information?
We take the security of your data extremely seriously. All identifying company and user information is encrypted before being stored. Your passwords are securely hashed using the industry standard PBKDF2 hashing function. Your parameter information is stored in a dedicated database created just for your team. Your results are stored in a secure cloud storage container created solely for your team’s results. No user from outside your team, including Canopy itself, will see any of your data. No-one will even be able to see you are using our platform.
Can you model my car?
Our vehicle model implements only pure vehicle physics; all of the car-specific information is in the parameters, which are set by you, the user. Almost anything on a car which impacts the dynamics is represented in our model, and if required we can extend the model on request, even making those extensions exclusive to your team if desired. In short, if your car has four wheels it can be faithfully modeled on the Canopy Platform.
Is your vehicle model suitable for a driver-in-the-loop simulator?
Yes, our vehicle model is a fully dynamic model which is more than capable of running in a driver-in-the-loop simulator at 1kHz which is the gold standard in professional motorsports. Running our vehicle model in your driver-in-the-loop simulator alongside your Canopy Platform subscription will give you an unparalleled ability to correlate your model, develop your car, and get feedback from your driver – all at record speed and with perfect alignment between offline and live simulations. We aim to interface with any brand of hardware or graphics/world model and support Dynisma, Ansible, Cruden, rFpro, vTAG, Speedgoat and more.