My First Driving Simulator - v1.0

I remember vividly the first discussion with some other PhD students back in 2005 regarding the nessecity to develop a platform that could test various automotive interfaces. Unfortunately such system or facility was non-existent in Scotland at that point. This was a major hinderance for the completion of my PhD as my proposed novel Human-Machine Interface for a Head-Up Display (HUD) should have been evaluated extensively and contrasted to existing Head Down Display interfaces. A driving simulator was deemed essential for the development and evaluation of the in-vehicle interfaces with particular interest to the proposed Head-Up Display (HUD) system. In particular the required driving simulator should be capable to recreate traffic conditions, accident scenarios and weather conditions whilst accommodating the HUD elements. However existing simulation suites and facilities were rendered unusable due to the projects financial constrains and the limitations of the existing simulators to accommodate the HUD interface.

In order to overcome this major issue, a more cost effective and customisable approach was employed through the utilisation of (licence-free) open source programs that can simulate various vehicles and driving conditions. These specialised simulators can be customised to create the ideal environment for testing various driving scenarios. Nonetheless, as the majority of these software solutions have been developed for gaming purposes (racing), they require substantial changes to the core of the programme in order to comply with real driving scenarios as required in order to evaluate the effectiveness of the proposed HUD systems. Particular emphasis was given in the vehicles’ AI which should recreate a realistic driving environment [1]. Conversely, a combination of off-the-shelf hardware (PC) and open source driving simulation software was the most cost-efficient and flexible solution.

Figure 1: Screenshot of Driving Simulator v1.0 based on Open source software TORCS

The idea of a custom driving simulator project was the initiative that brought together a multidiscipline group of researchers from different European and American universities. Through this project I had the unique experience to lead and coordinate a number of researchers operating form different places across the globe.Notably the members of that team completed the different sections of the software remotely, hence some of them they never met each other in real life [2, 3].

The functionality of the HUD interface components have been transferred to the simulated environment as it would purportedly be in a real vehicle. The symbols were programmed to follow the movement of the vehicles and alter their size and colour according to the relevant distance from the user’s vehicle. The two dimensional projection of the driving simulator did not obviously offer depth of field, which could have further enhanced the realism of the experiment. However, the advanced AI of the robot vehicles and the accurate superimposition of the HUD symbols enhanced the sense of presence; thus the derivable results were based on genuine driving efforts to respond to the accident events [3].

Given that the primary aim of the study was to measure and compare drivers’ response times, with and without the proposed HUD system, in contrary weather conditions, such conditions were replicated accordingly. Hence heavy fog, with visibility less than 50m, was simulated in four of the driving scenarios. The visibility percentage could be adjusted through the initial software to achieve a satisfactory fog quality. Specific parts of the code were improved in order to simulate the desired fog conditions and realistic depth of field. The developed simulator (v 1.0) in action is shown in the above figures.

Selected Publications Related to this project:

1. V. Charissis, and S. Papanastasiou, (2008), Artificial Intelligence Rationale for Autonomous Vehicle Agents Behaviour in Driving Simulation Environment, bookchapter in the “Robotics, Automation and Control” I-Tech Education and Publishing KG, Vienna, Austria, EU. (Bookchapter).

2. V. Charissis, S. Papanastasiou, and G. Vlachos, (2008), Comparative Study of Prototype Automotive HUD vs. HDD: Collision Avoidance Simulation and Results, in Proceedings of the Society of Automotive Engineers World Congress 2008, 14-17 April, Detroit, Michigan, USA. (SAE Best Presentation Award)

3. V. Charissis, S. Arafat, W. Chan and C. Christomanos, (2006), Driving Simulator for Head-Up Display Evaluation: Driver’s Response Time on Accident Simulation Cases, Driving Simulator Conference, Asia /Pacific DSC’06, Advanced Institute of Science and Technology, (AIST), Tsukuba/Tokyo, Japan.