At the Finish Line

It turned out that driving the Samara took some practice and was harder than steering a conventional car. However, it furnished a fair approximation of a natural driving situation and offered the physical environment of a real car interior. Our subjects were given the chance to get used to the simulator before they took up actual tasks. They were asked to find an individual driving speed that was as fast as they could go while still feeling in total control of the Samara, and then to maintain that speed throughout the tests. Some people had trouble getting used to the car, but most projected themselves into the driver’s seat fairly easily. The subjects generally avoided collisions and were clearly sorry if they did crash. (The collision sounds in the game were particularly impressive.) After the test concluded, some people parked the car on the side of the road so that other traffic could pass.

Because of the unrealistic nature of the simulator, there was no intention to make any absolute measurements. All tests were done as benchmark tests with several new solutions compared to each other, or with one new solution compared to existing solutions.

Still, the Samara was not just a test environment. At least as important was the fact that the designers themselves could step into the Samara and try the different solutions. In this respect, the Samara functioned better than more sophisticated driving simulators. We now believe that the design environment must be available to the design team for a drive at any time to try a new solution or to recall attributes about old ones. This kind of ready availability is impossible with the expensive simulators; they cannot be reserved for one project’s use for a long enough time, and they frown on drilling new holes into the design environment or messing it up with hot glue and tape. If an environment is going to accommodate new ideas, it must be open to modification.

An interactive prototype—demo, simulation—is the core of a UI design. It plays an important role in planning, testing, and communication. It forces a detailed specification of the interaction, enables the user’s active participation in the process, and helps communicate ideas forward. Testing the simulated concepts is an essential part of the prototyping process, and the methods used in it evolve and take shape alongside the available concretizations.

There were plenty of shortcomings with the Samara. Measurements were not as controlled as they should have been. Drivers might have compensated their performances by adjusting their driving speed in ways that were difficult to take into consideration when analyzing the results. The game caused random incidents that might have influenced the results. The driving feel could have been improved so that the performance challenge corresponded more closely to real driving situations. Automatic recording of performances would have saved a lot of time in the results analysis phase. The simulation environment should have allowed for measuring drivers’ reactions to a car braking in front of them; this was not possible with the Samara. The list goes on and on.

Nevertheless, the Samara fulfilled the expectations set for it. We were able to use it for preliminary testing, which was the target level in the Steering Wheel project. (Others took the ideas that seemed worth additional attention and continued to explore them.) Samara has also been used in other tests when the goal has been to research the use of a phone as a secondary task. It has been easy to modify; there have been numerous versions of the dashboard alone.

The Samara was cheap. All pieces of equipment in the simulation environment were intended for reuse. They are in fact still used in other research projects. Only the chassis and the PlayStation controls belong solely to the Samara. In the Steering Wheel project, the flimsy wreck of a Russian-made car with its tangle of wires, tape, and cardboard also functioned as a symbol of focusing on the essential.

Everything around me is technology—computer, projector, speakers, telephone exchange, cables, displays, and air-conditioning equipment. I am sitting in a dim room, inside a metal chassis that was once a car. In principle, the simulation is ready, but all the task paths have to be tested and the probable error scenarios must be logged. It was already late when I started, but I have to be methodical and thorough. I start up the simulation time after time, always recording the adjustments. The simulation incorporates an audio system. Every time it starts up I hear its CD player repeating how we have the right to be carefree, and how easy money and life make you laugh. There are still things to be checked out.