Determining the allocation of tasks in automated high speed train driving

Abstract

Adhering to the increasing demands in terms of numbers of passengers, travel time, energy efficiency and optimal exploitation of infrastructure, from an operational point of view future high speed trains are in need to evolve towards higher grades of automation (GoA). Nowadays higher grades of automation (GoA2 and beyond) are mainly found in subway systems or other specialized systems that basically provide a closed system environment. Introducing higher grades of automation into high speed train operation covering long distances above ground poses significant challenges to safe and secure operation. The train driver has always played a major role in railway safety, but he is increasingly assisted by technical components. Introducing higher grades of automation to high speed train operation is seen as a step further along the path of pairing up technological assistance and human performance to reach a maximal level of safety. The role of the train driver within a highly automated system needs to be considered in detail in order to provide clarification what tasks the future train driver is able and needs to execute reliably to ensure the required level of operational safety. To this end a series of studies is set into place at the German Aerospace Center e.V. (DLR) embarking with semi-structured expert interviews, later involving scenarios being tested in a high fidelity simulator and finally participatory design approaches being applied to prototype potential cabin design measures based on empirical research. The current paper describes the first step taken in this current line of research. Based on an interview guideline covering 29 questions, two hour semi structured interviews with five high speed train drivers and trainers touched upon the current state and possible future development of in cab signaling, informational sources, and function allocation in two scenarios. One scenario included a braking procedure in a regular operational environment. The other scenario was concerned with system failure and measures to restore safe operation. The possible future development of the tasks and means were based on certain parameters (GoA 2, an operating speed of 400 km/h, the moving block logic, and a reduced outside view). Additionally, the experts indicated their view on future task allocation for ten specific tasks provided on a continuous dimension ranging from train driver to technical assistance. From this, a classification of the ten specific tasks regarding the function allocation between human and machine was generated. Based on this classification and an earlier developed HTA (Hierarchical Task Analysis) of the current tasks of a high speed train driver, the results of the interviews were integrated into a comprehensive set of future tasks. This set of future tasks will be thoroughly discussed in the paper. Additionally, the interviews gave first hints where technical assistance in the cabin needs to be implemented to ensure safe operation, e.g. in terms of a technically enhanced preview of the route ahead.