Find Us On Social Media
Tag CloudADAS ANPR Autonomous Vehicles Brake Cameras CIS Collision Avoidance Connected Vehicles cycling DFT Distracted Driving Drink Driving Driver Behaviour Driver Monitoring Enforcement EU Events Fleet Management FTA Government GPS HGV IAM Insurance ITS Legislation LPR Pedestrians Police PSV reports road construction Russia security Statistics Surveillance Telematics Traffic Management Tyres UK USA Vehicle Operation Vehicle Technology Vehicle Testing Volvo
New platform for advanced auto engineering simulation
Jonathan Newell tries out the Delta Series simulator from Ansible Motion for vehicle dynamics and ADAS testing and development.
Located at the modern Hethel Engineering Centre near Norwich, Ansible Motion’s high technology facility is poised to expand from its motor racing heritage into the development laboratories of the mainstream global automotive manufacturers.
To do so, it has the significant challenge of overcoming the competition in an industry that has become so familiar with its large, ungainly “hexapod” simulation platforms, that you could almost say it’s entrenched.
However, as the company’s Technical Director, Kia Cammaerts explained, it’s the inability of the historic hexapods to meet the demands of modern vehicle dynamics that puts Ansible Motion’s new platform at an advantage.
From Hexapod to Stratiform
Hexapods are familiar to all as the matrix of computer controlled rams on which a simulator sits which provides the motion. However, the complex motion needed by the rams to achieve even simple simulator motion adds latency and inertia into the system.
The reaction lag of such simulators may be fine for entertainment or for simulating ships or aircraft where pilot input is based on long visual horizons and measured input, but has been found to be increasingly unsuitable for the short reaction times and highly dynamic environments of motor vehicles.
This was particularly the case in the motor racing industry, where simulation latency made it impossible to adequately model racing car dynamics.
As a result, Ansible Motion set out to redesign the motion platform entirely and came up with the stratiform model, comprising an X-Y stage for front-to-back and side-to-side motions. The table on the stage also provides yaw, pitch, roll and bounce motions, coupled with internal simulator pod motions, which in total provide the pod with over 11 motion axes with much less inertia than a hexapod.
Advanced software modelling
Putting the entire simulation package together requires the co-ordinated integration of motion control, parameterised physics models and environmental models.
The physics models can be synchronised directly from such tools as Dassault Systèmes’ Simpack and define the vehicle parameters such as suspension damping, roll control systems or driver assist systems (ADAS). The environment models, such as are available from rFpro define the roadway, pot-holes, rumble strips and other aspects of the driving environment.
The amount of data that needs to be processed is vast and so the potential for latency (or lag) due to processing power is as much of a factor as the mechanical lag of high inertia systems such as the hexapod. To overcome this, Ansible Motion uses its own custom-built array of sixteen computers with parallel processing.
To fully understand vehicle dynamics and the way a car feels on the road, as well as making adequate assessments of the latest Advanced Driver Assist Systems, it’s important to put a driver into the simulation environment to see how real drivers interact with vehicle systems.
Using a combination of Driver-in-the-Loop (DIL) and Hardware-in-the-Loop (HIL) modelling provides the best combination for making assessments of vehicle behaviour based on driver behaviour or driver reaction to active safety systems.
The difficulty with driver-based simulation systems is making the experience as real as possible and ensuring that every driver input is evaluated in the model and given an appropriate system response.
Having been developed for the racing industry with expert drivers, the attention to detail needed to be extremely fine so that even the tiniest cue to the onset of drift, for example, could be felt by the driver and responded to in an appropriate way.
There are a number of such cues within the simulation toolbox including environmental, contact and inertial. Kia Cammaerts explained to me that the inertial cues are what provide the driver with the feeling of motion and if these don’t match the visual cues, motion sickness will be the result.
On the track
As I sat in the simulator pod and the lid was lowered and the platform moved into its neutral position to start the simulation, the overall feel was one of a piece of advanced engineering equipment designed with the professional driver in mind – this was certainly no amusement park attraction.
The 8-metre diameter wrap round screen illuminated and I was immersed in the starting grid of the Spa Francorchamps circuit. The software had provided me with a standard saloon car rather than the kind of high performance racing car that would be impossible to handle so the Spa circuit felt more like a white-knuckle drive through the Belgian countryside than an all-out battle to reduce lap times.
It was clear from the outset that the all-important cueing was perfect. Every bump, rumble and handling limitation was felt despite not moving outside the small radius of the stratiform platform.
Slamming on the brakes approaching a corner at too high a speed results in the simulator giving your vestibular system a virtual prod so you feel the onset of harsh braking. In my case, it was too late and a flurry of cues made it quite clear I’d flown over the rumble strip and was spinning on the grass.
For me, driving the Delta Series simulator from Ansible Motion had been an interesting experience as a driver and as an engineer but for those in the profession, this new platform is enabling the development and prototyping of advanced vehicle dynamics and ADAS technology that will bring autonomy and vehicle safety to our roads much sooner than otherwise could have been hoped.
Studied Engineering at Loughborough University and now involved in broadcast and technical journalism. Jonathan is based in London and Almaty.
Latest posts by Jonathan Newell (see all)
- New platform for advanced auto engineering simulation - Tuesday, 31 May 2016
- FBI warns of cyber threats to connected vehicles - Tuesday, 22 March 2016
- Putting intelligent mobility through its paces - Tuesday, 1 March 2016