On “Wicked” Problems in the Design of Autonomous Vehicles Brigitte Jordan
Download 17.6 Kb.
|
| On “Wicked” Problems in the Design of Autonomous Vehicles
Brigitte Jordan Last changes: Monday, April 20, 2015 by gj keywords: ADAS, AV Major Issues for the Design of Autonomous Vehicles Humans and their technologies have been co-evolving for thousands (and probably millions) of years. As a consequence, any significant new tool tends to generate significant changes in the tool users’ physiology, cognition, and social relationships as has been amply demonstrated by archeologists and other paleo-scientists. In the past, this took place over long periods of time, but more recently the digital revolution has pushed forward tool and car design in ways to which the human body and mind have not yet adjusted. We are now faced with the transition from driving in a conventional, level-1 car to a level-4 full-AV vehicle - two modes of transportation that are fundamentally different. It may be the case that the mental and physiological adjustments that could make that transition smooth will take too long a time to occur. Nor are there any obvious ways to hasten that transition, either on the human or the vehicle side. This gap between the capabilities commanded by a human driver and those programmed into the automated car is a major issue at the core of what is required to bring fully automated vehicles into reality. It represents a gulf across which it is difficult if not impossible to jump. Handing over control There clearly will be situations, especially before all transport vehicles are automated and all roads are properly mapped and improved to handle AV vehicles, when it becomes desirable or necessary to hand over control, be it from a driver to the machine or from a machine to a human operator. When control has to be handed from automated car to human driver, the issue is particularly problematic since this driver may not have driven a car for years. S/he may also be incapacitated, could be playing games at the moment, or be involved in new kinds of activities made possible by advanced forms of ASAD that will have become available. Without the continuous updating of skills that happens in an active driver, the question arises, does this person still know how to drive? especially in an emergency? Why have designers consistently acknowledged and then ignored the massivity of the hand-off problem, favoring a gradual approach that allows them to put their faith into ADAS-type small improvements but ignore the fundamental issue, which is that there is no way that we can effectively counteract the deterioration of the (former) driver’s skills in the face of the increasing sophistication of the car. No amount of training or of warning signals can remedy this kind of unremediable mismatch. Sometimes suggestions are mad that the cognitively absent driver could be alerted or should be trained, but this is a “wicked” problem for which our current thinking yields no comfort and no solution. It is not solvable from within the existing system. Wicked Problems The idea of “wicked problems” was introduced decades ago by organizational scientists from UC Berkeley to describe problems that are difficult or impossible to solve because of their formidable complexity and the entanglements of their multitudinous variables (Rittel and Webber 1973). Wicked problems, they found, have no precise formulation and are of a scope that makes them unique, so no leverage is gained by looking for insights from comparison. Their solution, if one occurs, never has a scientific test. The concept of wicked problems made clear how inadequate a sequential, structured methodology is for understanding complex design problems. Wicked problems cannot be successfully treated with traditional linear, analytical (systems-engineering-like) approaches (Frankel and Racine 2010) because what drives them are fundamentally not technical but social or cultural problems. What is to be done? Solutions or paths to solutions for the wicked problem are murky. What is certain is that the solution will not lie in simply transferring control from a disabled driverless robot car to a human driver. Nevertheless, we would suggest that what situations like the hand-off problem require is to abandon the current logic, a logic that sees a driver (human or machine) in control, and to step outside the box, abandoning the image of driver inside car, even the popular option of car and driver as a team. The way out may be to turn to the world beyond the disabled robot car and its fate mates and position it together with its driver, its occupants and its connections in a wider environment. As the poet Jane Hirshfield tells us, “Everything is connected. Everything interacts with everything else”. And she adds: ”Pay attention.” Little can be learned about a wicked problem by structured data gathering and analysis. With the situation in constant flux (Ken Anderson and the Intel crowd) wicked problems demand an opportunity-driven approach; they require making decisions, doing experiments, launching pilot programs, testing prototypes, and so on. They need to accommodate multiple alternative perspectives rather than foresee single solutions. This suggests that we might employ the ethnographic methods of design anthropology to gain a deeper understanding of the resources that might be available inside and outside the affected car, be that through automatic connection to other vehicles or initiatives taken by humans in the car behind or on its side. In any case, the problem requires a larger scope, a wider solution space. References Buchanan, R. 1992 Wicked Problems in Design Thinking. Design Issues 8:2 5-21 Frankel, Lois and Martin Racine 2010 The Complex Field of Research: for Design, through Design, and about Design. Proceedings of the 2010 Design Research Society International Conference. Montreal, July 7-9 Rittel, Horst and Melvin Webber 1973 Dilemmas in a General Theory of Planning. Policy Sciences 4, 155-169. Footnotes re evolution In order to do that, we would need to close the gap between the two modes of driving (conventional human vs. a robotic, fully automated car). What works against that is that the trend runs in the opposite direction. Cars more and more are able to handle their growth in automation while humans are more and more involved in happily loosening themselves from the shackles of constricting requirements. re “connected” p3 Jane Hirshfield in an interview with Elizabeth Lund, of The Christian Science Monitor Everything is connected, everything changes. Pay attention.
Frankel and Racine called these wicked problems, in contrast to tame problems that have well defined and stable problem statements. Wicked problem theory presented an alternative to the scientific approach that had been espoused by design research. Pieces and leftovers Frankel and Racine: Rittel and Webber (originated) their concept of wicked problems, which made clear, among other things, how inadequate a sequential structured methodology was for understanding complex design problems. Characteristics include the impossibility of formulating an exact problem due to the complexity of evolving variables, a situation ethnographic interpretive methods are best able to handle. Frankel and Racine pointed out that wp’s cannot be successfully treated with traditional linear, analytical (systems-engineering-like) approaches. Directory: Papers Papers -> From Warfighters to Crimefighters: The Origins of Domestic Police Militarization Papers -> The Tragedy of Overfishing and Possible Solutions Stephanie Bellotti Papers -> Prospects for Basic Income in Developing Countries: a comparative Analysis of Welfare Regimes in the South Papers -> Weather regime transitions and the interannual variability of the North Atlantic Oscillation. Part I: a likely connection Papers -> Fast Truncated Multiplication for Cryptographic Applications Papers -> Reflections on the Industrial Revolution in Britain: William Blake and J. M. W. Turner Papers -> This is the first tpb on this product Papers -> Basic aspects of hurricanes for technology faculty in the United States Papers -> Title Software based Remote Attestation: measuring integrity of user applications and kernels Authors Download 17.6 Kb. Share with your friends:
|