Mascha C. van der Voort

COMBINING KOHONEN MAPS WITH ARIMA TIME SERIES MODELS TO FORECAST TRAFFIC FLOW

A hybrid method of short-term traffic forecasting is introduced; the KARIMA method. The technique uses a Kohonen self-organizing map as an initial classifier; each class has an individually tuned ARIMA model associated with it. Using a Kohonen map which is hexagonal in layout eases the problem of defining the classes. The explicit separation of the tasks of classification and functional approximation greatly improves forecasting performance compared to either a single ARIMA model or a backpropagation neural network. The model is demonstrated by producing forecasts of traffic flow, at horizons of half an hour and an hour, for a French motorway. Performance is similar to that exhibited by other layered models, but the number of classes needed is much smaller (typically between two and four). Because the number of classes is small, it is concluded that the algorithm could be easily retrained in order to track long-term changes in traffic flow and should also prove to be readily transferrable.

A prototype fuel-efficiency support tool

An effective way to reduce fuel consumption in the short run is to induce a change in driver behaviour. If drivers are prepared to change their driving habits they can complete the same journeys within similar travel times, but using significantly less fuel. In this paper, a prototype fuel-efficiency support tool is presented which helps drivers make the necessary behavioural adjustments. The support tool includes a normative model that back-calculates the minimal fuel consumption for manoeuvres carried out. If actual fuel consumption deviates from this optimum, the support tool presents advice to the driver on how to change his or her behaviour. To take account of the temporal nature of the driving task, advice is generated at two levels: tactical and strategic. Evaluation of the new support tool by means of a driving simulator experiment revealed that drivers were able to reduce overall fuel consumption by 16% compared with ‘normal driving’. The same drivers were only able to achieve a reduction of 9% when asked to drive fuel efficiently without support; thus, the tool gave an additional reduction of 7%. Within a simulated urban environment, the additional reduction yielded by the support tool rose to 14%. The new support tool was also evaluated with regard to secondary effects.

The influence of time-criticality on Situation Awareness when retrieving human control after automated driving

When applying automated driving as a means for congestion assistance, developers are challenged how to accommodate the transitions between automated and manually driving, especially because these transitions might occur regularly and suddenly. During automated driving, the ability to take over control is also aggravated due to the driver being placed out of the control-loop. To assess then the ability to retrieve human control, we tested within a driver simulator experiment the influence of criticality (available time) on Situation Awareness (SA) gained during time-critical take-overs within a scenario of congested driving. Though one of the applied measurement methods did not show the expected effect of SA on successfulness of taking back control, the results show that drivers are able to successfully retrieve control, also within time-critical situations. Furthermore, the results show that the ability to retrieve control is positively influenced if drivers gain increased levels of SA.

Retrieving human control after situations of automated driving: HOw to meaure Situation Awareness

When applying automated driving as a means for congestion assistance, developers need to account for regular and sudden transitions between automation and manual control. To assess then the ability to retrieve human control, we tested within a driver simulator experiment how Situation Awareness (SA) gained during time-critical take-overs, and based on a simulation freeze technique (SAGAT) and SA queries (SART), could be measured. The results show that drivers are able to build up SA in these time-critical situations. With regard to measurement method, only SART showed positive correlation with successfulness in retrieving control. Nonetheless, the applied measurement method showed to allow future selection of support types drivers are in need for when retrieving control after automation.

Expanding the representation of user activities

Space is often designed based on the representations of user activities (i.e. lists, organograms or flowcharts) that streamline user activities in stepwise, oversimplified, representations that may leave insufficient room for future activity development. However, design can anticipate activity development if users are able to represent their own activities while participating in the design process. A case study of a medical imaging centre reveals that once users have such opportunity, their spatial practices are not only taken into account but also expanded. The designers, the users and the researchers created a range of instruments to expand across three units of analysis: operations, actions and activities. As a result, the representations of space proffered by the designers were expanded to a space of representation for the users, where new ways of working were realized. Based on this study, an integrated model for the production of space and the development of activity is proposed.

User-centred system design approach applied on a robotic flexible endoscope

Abstract Complex systems, like surgical robots, are designed by engineers. It is very difficult for them to determine the different needs and desires of all stakeholders. Especially when designed from scratch, end user input is essential in creating a system that has added value, is user friendly, and can be easily integrated into practice. For the development of a robotic flexible endoscope we have involved physicians, nurses, and equipment suppliers in our design approach. Seven steps are executed to convert user preferences and capabilities into concepts: • Determine focus area of development. • Create the current workflow of system application to understand (the context of) use. • Determine problem definition and design goal. • Create the future workflow, in which current problems are eliminated and major system wishes are fulfilled. • Translate the future workflow into a functional overview that contains system functions. • Select and configure the appropriate construction elements into physical overviews, being preliminary concepts. • Decompose physical overview into manageable modules. These views are evaluated by the major stakeholders and together form a system architecture. The system architecture helped us in defining the robotic modules required to fulfill all stakeholders‟ needs and desires. Demonstrators were built to evaluate critical concepts in clinical relevant experiments.

Implementing Human Factors within the Design Process of Advanced Driver Assistance Systems (ADAS)

This paper introduces our research which aims to develop a design approach for ADAS applications in which human factors (including stakeholder feedback and objective performance measures) are explicitly accounted for. Since driving is part of a complex (traffic) system, with a large number of interacting components, ADAS design is confronted with choices for which the influence on the system, and the driving performance in particular, is not immediately manifested. Therefore, providing designers with relevant feedback during the design process, about the consequences of specific choices, will increase the efficiency and safety of driver assistance systems.

How to assess driver's interaction with partially automated driving systems - A framework for early concept assessment

The introduction of partially automated driving systems changes the driving task into supervising the automation with an occasional need to intervene. To develop interface solutions that adequately support drivers in this new role, this study proposes and evaluates an assessment framework that allows designers to evaluate driver-support within relevant real-world scenarios. Aspects identified as requiring assessment in terms of driver-support within the proposed framework are Accident Avoidance, gained Situation Awareness (SA) and Concept Acceptance. Measurement techniques selected to operationalise these aspects and the associated framework are pilot-tested with twenty-four participants in a driving simulator experiment. The objective of the test is to determine the reliability of the applied measurements for the assessment of the framework and whether the proposed framework is effective in predicting the level of support offered by the concepts. Based on the congruency between measurement scores produced in the test and scores with predefined differences in concept-support, this study demonstrates the frameworks reliability. A remaining concern is the frameworks weak sensitivity to small differences in offered support. The article concludes that applying the framework is especially advantageous for evaluating early design phases and can successfully contribute to the efficient development of drivers in-control and safe means of operating partially automated vehicles.

Participants' interpretations of PD workshop results

While the majority of participatory design (PD) research deals with the development of tools and the analysis of occurrences during workshops, we believe more research into the meaning and value of PD study results is needed to make PD more attractive to design practice. In this paper we present how forty participants of an ongoing one-year healthcare project interpret the outcomes of participatory design sessions with varying topics differently and what can be learned from that. We find that different workshop topics and different roles in the project lead to significant differences in how workshop outcomes are interpreted by individual participants, project managers, and researchers. These results illustrate the value of an inventory of the interpretations of individual participants of workshops results so that we are able to dynamically adjust successive sessions to match these perspectives and hence improve the success of the whole PD study.

Human-centered challenges and contributions for the implementation of automated driving

Automated driving is expected to increase safety and efficiency of road transport. With regard to the implementation of automated driving, we observed that those aspects which need to be further developed especially relate to human capabilities. Based on this observation and the understanding that automation will most likely be applied in terms of partially automated driving, we distinguished 2 major challenges for the implementation of partially automated driving: (1) Defining appropriate levels of automation, and; (2) Developing appropriate transitions between manual control and automation. The Assisted Driver Model has provided a framework for the first challenge, because this model recommends levels of automation dependent on traffic situations. To conclude, this research also provided brief directions on the second challenge, i.e. solutions how to accommodate drivers with partially automation