Agent-based Human-computer-interaction for Real-time Monitoring Systems in the Trucking Industry

Auto ID systems can replace time-consuming, costly and error-prone processes of human data entry and produce detailed real time information. However, they would add value only to the extent that data is presented in a user-friendly manner. As model-based decision support is not always adequate, an agent-based approach is often chosen. Real life entities such as orders and trucks are represented by agents, which negotiate in order to solve planning problems. For the respective data representation at least two forms can be distinguished, focusing either on (1) resources (account-based) or (2) orders (order-centric). Applying cognitive fit theory we describe how the different interfaces affect decision making. The hypotheses would be tested in a laboratory experiment. The intended contribution should support that order-centric interfaces have higher user-friendliness and are especially beneficial to low-analytics and planners working under time pressure