This is the first Driver in the Loop Simulator at Enstone

Lotus Renault are set to add a new simulator facility by the beginning of the 2012 race season in March. Here, James Allison tells us more…

Why do teams have simulators?


JA: Teams have had simulators for decades. There are many types of simulator, and they are used for many purposes. For example, a team might have a hydraulic loading rig which is capable of subjecting a wishbone to a series of loads that accurately simulates the burden it will carry when racing for real - in this case, the simulator is a way of ensuring in a controlled manner that the bone will perform safely when you send it racing.

James Allison Technical Director, Lotus Renault GP
James Allison Technical Director, Lotus Renault GP

Photo by: xpb.cc

A team might have a different form of simulator where the behaviour of the aerodynamics, the suspension, the tyres, the gearbox and the engine are all accurately represented as a series of equations in a computer routine. The same computer would have a rudimentary mathematical model of a driver that attempts to "drive" the virtual car around a virtual circuit in the computer to return information about the behaviour of the car around the lap.

This type of simulator (often referred to as a "Lap Simulator"), which has been in use for over quarter of a century is a vital part of preparing and optimising the setup of the car prior to arriving at a race - for example, in this virtual environment, items such as spring and wing settings may be tried out and the ones which produce the best simulated laptime retained.

The type of simulator that this article is referring to is a Driver in the Loop simulator. In many ways, such a simulator is similar to Lap Simulator described above. The key difference is that in place of a rudimentary mathematical model, a real driver provides the control inputs to drive a virtual car around a virtual lap.

Although there are a host of problems with using a real driver in place of a mathematical driver model, the real driver brings a capability to the simulation that cannot be matched by a computer model. The reasons for this are complex, but a simple explanation for this is as follows: It is not yet well understood precisely how a racing driver controls a racing car when the car is close to the limit of tyre grip.

A computer is capable of driving an unstable virtual car in a manner that the human driver would not be capable of. This difference between real driver and computer model driver leads the Lap Simulator approach to make serious errors in its recommended setups whenever the engineers are trying to assess changes in the driveability of the car.

The Loop Simulator is able to make setup recommendations

James Allison

By inserting a real driver in the simulation (the so called "driver in the loop" approach) a team is able to bypass the difficulty of providing an accurate mathematical model of a human and the Driver in the Loop Simulator is able to make setup recommendations to improve the car that could never have emerged from the Lap Simulator approach.

Will our new simulator replace an existing resource, or be a totally new asset?


JA: This is the first Driver in the Loop Simulator at Enstone, and the capability it will bring us will complement the suite of other forms of simulation that we have used for years.

What are the factors for consideration when creating a simulator?


JA: Like the Lap simulator, there must be a mathematical model that accurately describes the physics of the real vehicle. Creating this is a daunting undertaking, but the underlying physics model is common with existing simulation methods. The areas of key difference to the Lap Simulator approach all involve the insertion of a real driver into the simulation loop.

The driver can only add value to the process if the simulation environment mimics reality with sufficient accuracy for the driver to feel like he is driving a real car. This means that the system must be very carefully designed to provide very accurate visual representations of the circuit and its environs. It must accurately reproduce the sound of the real car as this is one of the cues that the driver uses to judge things like braking and gear changes.

Finally, and most challengingly, it must move the driver in a manner that is convincing. It is not possible for a simulator to reproduce the G-loads that a real F1 driver experiences, but it is capable of reproducing particular aspects of the motion of the car that are very important for persuading the brain that the simulator is really driving around the circuit.

When the simulator is complete will our drivers then have to spend every spare moment hard at work on simulations?


JA: No. They will drive it, but it will be used for the majority of the working year, and it is not practical to use the race drivers in this manner. The main bulk of the work will be carried out by specialist simulator development drivers.

Is the simulator targeted at driver development or car development?


JA: It can perform both roles. A driver might use it for learning the layout of a new track, or for learning various procedural aspects of driving - the start procedure for example. However, the lion's share of the benefit will accrue in car development.