TRADR Team Communication Corpus

The TRADR Team Communication Corpus was created in the EU-funded project TRADR project (www.tradr-project.eu). In TRADR we investigated how humans and robots collaborate as a team in disaster response scenarios (such as an industrial accident or an earthquake). The main task of the human-robot team is to gather information for situation assessment: map the environment, detect victims, identify potential damages and hazards, collect samples during a mission. TRADR end-users are the human responders involved in a mission in various roles, including the mission commander, team leader, the robot operators or one or more infield rescuers. Robots are used as tele-operated tools or (semi-)autonomous team members to reach areas which are considered too dangerous to enter for humans.

The corpus contains audio recordings and transcriptions of human-team communication during missions carried out as part of TRADR Joint Exercises and Evaluation experiments. It is released under the Creative Commons license. It is free to share and adapt the datasets for non-commercial purposes in the way specified in the license terms (https://creativecommons.org/licenses/by-nc-sa/3.0/).

TJex 2015

From May 18th to 22nd 2015 the TRADR Joint Exercise took place at Phönix West, which is the ruin of an old furnace in Dortmund, Germany. The exercise was conducted in cooperation with the TRADR partners and firefighters from the Institut fur Feuerwehr und Rettungstechnologie Dortmund.

The TJex 2015 corpus consists of three interactions in German, and one in English.

TEval 2015

In September 2015 the TRADR Evaluation Exercise was held in Dortmund, Germany, at an old industrial complex of a former blast furnace, in collaboration with Dortmund firefighters.

The TEval 2015 corpus consists of four interactions in German.

TEval 2017

The last TRADR Evaluation exercise took place in the port of Rotterdam, the Netherlands, from 13 to 16 November 2017. The exercise was conducted in collaboration with Rotterdam Harbor United Firefighters.

The TEval 2017 corpus consists of six interactions in English.