The TRAINS project started in 1990 at the Computer Science Department of the University of Rochester. Led by Profs. James Allen and Len Schubert, the goal of the project was to combine two core areas of Artificial Intelligence research: natural language understanding and automated reasoning, in particular planning. There were predecessors for this combination, for example Winograd's famous SHRDLU program. But the TRAINS project was certainly among the first to combine spoken natural language conversation with intelligent decision support in the form of a problem solving assistant.

Based on human-human problem solving dialogues collected in our lab, the system presented simple transportation problems and helped the user solve them. It was used for a variety of research projects into spoken language interaction and intelligent decision support over the years, and eventually evolved into the TRIPS system.

This video, from 1995, provides an introduction to the project and overview of the system. For a recording of a complete session with the system, please see the next video in this playlist.

Contributors: Allen, Schubert, Martin, Traum, Poesio, Ferguson, Light, Miller, Heeman, Stent, ...