Cockpit

Dr Robert Wellington - Auckland University of Technology

At AUT we are conducting research into cockpit design that will inform design decisions of the physical JetBlack vehicle.  We have built a cockpit pod in our HCI (Human Computer Interaction) laboratory, where we can try different design ideas and collect data about driver interaction.  There are many sub-projects in this project, eg: designing and testing a customised yoke, and the design and incorporation of different visual display options being amongst the most complicated, and for us, the most interesting.

 

For us, building the 3D model and cobbling together plywood, dowel, fabric, and computer equipment means more than just prototyping.  It gives us an opportunity for researchers in multiple disciplines to work together and incorporate many of the strengths we have in the School.  The mathematical modelling, the generation of graphics, the creation of physical artefacts, and the collection of data about the actual use and performance of the combination of these things provide research opportunities across quite a range of research disciplines.

 

The range of research involved in this project range from the creation of an accurate (and realistic) physics model of the car and the environment to enable reliable simulation of different input and output methods and artefacts, through to the social, cultural, and psychological aspects of driving a vehicle in a high cognitive load environment.  For example how does the culture of flying a single seat fighter relate to the project and how does that translate into the procedures outside of the cockpit in terms of handing over to the driver to give them an assurance of control, and adequate information about status and performance in order to make appropriate decisions in the run?

 

There are many situations where the AUT team needs to work closely with both the driver and other systems engineers to design relevant research activities.  For example, determining how the braking system might be controlled is a non-trivial problem involving many decisions.  How many controls are required? Where might they be placed? Should the driver be able to over-ride the system?  The constraint of resources in terms of control surfaces and the limited physical space in the cockpit mean that the decisions involved in the braking system are related to decisions for the controls for propulsion etc.  In the end, as would be expected with such a project, nothing is simple when you want to design a vehicle to go this fast, but on the other hand, that’s what makes it so interesting for research.

 

What we have built so far starts with a hollow plywood and fibreglass base on casters to keep the weight down in case we need to move it. On the base we have built sides and a roll cage to match the physical space that will be available in the real car, this was built out of plywood and pine dowel. We have mounted a high end force feedback gaming steering wheel and pedal system in the car after making some modifications. We had earlier built a rudimentary HUD (heads up display) that did work but wasn’t optimal, so we are going back to the drawing board to look at the materials and reflective coatings, along with the best strategy for projection – at this point there is very little space to play with to fit this in, but we believe it is possible and are currently using an old tablet PC to stand in for the HUD (although you obviously can’t see through it).  The Tablet computer picks up data from the main computer over the network and we can configure this to display virtually any data produced in the simulation, at present we have status data for the rocket thrusters, main thrust, and speed.  We have a home theatre projector to display the world view, and everything is plugged into a gaming PC with good graphics processing. The world and the vehicle are modelled in a game development application and this allows us to apply forces to the vehicle that are related to the interaction between the vehicle and the world, and respond to driver input.

 

So far we have developed a simulation that is subsonic, and it does feel pretty real and requires real skill and concentration to have a successful run, and so probably represents the real challenge at these speeds.  Our ongoing task is to collaborate with the other systems designers to fine tune the various ‘numbers’, things such as; drag coefficients, force applied by the wheels for steering at varying speeds, and obviously thrust forces.  We will be printing a custom yoke in AUT’s rapid prototyping lab (yes – that is a 3D printing process, but that’s another story) and working through the interaction issues.  All the while we will be getting participants to use the simulation, video them, and learn about the design issues – perhaps have a few goes ourselves.  It’s hard work, but someone has to do it.