Navigation Bar HOMECONTACTPROJECTSPUBLICATIONSRESOURCESSEARCHSTAFFface and bodyhuman motionpheromonessimulationurbanisation

DEPARTMENT OF ANTHROPOLOGY- HUMAN BEHAVIOR RESEARCH | human motion

 

logo

CADDY - Cognitive Autonomous Diving budDY

 

Project summary


The goal of the caddy project was to develop an autonomous underwater robot who was able to assist a diver, interact with the diver and control the divers state. The technical challenges for our research were the development of diverNet - a motion capturing array of sensors and devices for recording of heart  rate and diver breathing. Another challenges was the devlopment  of an underwater tablet which could be used by the divers to communicate and fill out questionnaires.




We developed multilayer perceptrons for diver behaviour interpretation using emotion, motion and physiological data collected in dry land and diver experiments.
We showed as a proof of concept that it is possible to determine experienced emotional states from body motions captured with DiverNet motion sensors.
Although the information is present in breathing motions – the primary source which give
s much better results are body motions. Using body motions for prediction is not only possible in walking behaviour but also for a diver swimming underwater. Thus produced motion is a basic predictor for experienced emotioal states in  a pleasure arousal and dominance circumplex model. As far as we know there is no scientific report on this relation up to now.  Another highlight is the categorization of behaviour categories from body motions. A neural network can recognize what the diver is doing at the moment. In addition we developed a real time analysis program which plays back dive recordings and analyses them in real time. Such systems are of gret interest for mission planning and mission surveillance, for instance in SWAT teams or in space when visibility is low or zero.

See the final results here.

Divernet sensor array

DiverNet


DiverNetControl Center

Divernet Control Center




Caddy was funded by the European Community's Seventh Framework Programme under Grant Number FP7-ICT-2013-10-611373

CONSORTIUM

  • SVEUCILISTE U ZAGREBU FAKULTET ELEKTROTEHNIKE I RACUNARSTVA Croatia
  • CONSIGLIO NAZIONALE DELLE RICERCHE Italy
  • INSTITUTO SUPERIOR TECNICO Portugal
  • JACOBS UNIVERSITY BREMEN GGMBH Germany
  • UNIVERSITAET WIEN Austria
  • UNIVERSITY OF NEWCASTLE UPON TYNE United Kingdom
  • Divers Alert Network Europe Foundation DAN Europe Malta
 

UNIVERSITY OF VIENNA

all rights reserved karl.grammer@univie.ac.at