When the control centre thinks for itself: human-centred AI for public transport : Date:
Public transport control centres are receiving an everincreasing volume of information. The IADAPT research project at Furtwangen University is developing adaptive user interfaces that adjust to traffic conditions, the task at hand and the control centre staff. The focus is on human-centred AI, which does not replace people but provides them with transparent support when making complex decisions.
When a football match ends, a stretch of road is closed or several vehicles break down at the same time, control centres have to assess a vast amount of information in a short period of time. At the same time, the number of mobility services and data sources is growing. Static user interfaces reach their limits when they display too many notifications at once – regardless of what is actually relevant to the task at hand.
The Institute for Intelligent Interactive Ubiquitous Systems (IIIUS) at Furtwangen University is developing a new generation of interactive control centre systems for public transport as part of the IADAPT research project (Intelligent User Interfaces for Adaptive Control Centre Systems), in collaboration with the HFU Tuttlingen Campus and the industry partner INIT. A key milestone is the prototype recently set up in the laboratory: it provides the basis for testing and further developing adaptive operating concepts and interaction strategies for the control centre of the future under realistic conditions. The aim is to create a control centre that can dynamically adapt to the current situation, the task at hand and the needs of the control centre staff. IDAPT is funded by the Federal Ministry of Research, Technology and Space (BMFTR) under the ‘FH-Kooperativ’ funding scheme.
The system is designed to identify which information is relevant in a specific situation, how it should be presented in a clear and understandable way, and what interaction options are required. In the event of a fault, for example, different information takes precedence than during normal operations. The system also takes into account collaboration within the control centre team and communication with other parties involved.
Human-centred AI as a collaborative development process involving people, organisations and technology
How can intelligent systems effectively support people without taking control away from them or making decisions opaque? In safety-critical working environments in particular, technical capability alone is not enough. IADAPT therefore consistently combines technical development with user-centred methods. Work processes, information flows and existing user interfaces are analysed; requirements and interaction strategies are developed collaboratively with a focus on practical application. At the same time, the project investigates how traditional methods of user-centred design can be applied to systems whose behaviour and presentation change dynamically. Criteria such as effectiveness, efficiency, comprehensibility, trust and satisfaction are thus integrated into system development from the outset.
The technical basis is a flexible, event-driven software architecture. It links different data sources and interface modules and adapts the display according to the situation. This modular approach can also be applied to control centres in the energy supply, manufacturing, logistics and civil protection sectors.
As a practical partner, INIT contributes its Intermodal Transport Control System as well as experience gained from real-world traffic management systems. This fosters close collaboration between applied research and industrial development. The results are intended to help make processes in control centres clearer and more efficient, relieve staff of the burden of complex decision-making, and strengthen the resilience of public transport.
Text source: Furtwangen University