8 July 2013 - EUCog "Message Meeting", Munich

Final Report

Brief Report on the EUCog "Message Meeting", Munich
July 8th, 2013
Vincent C. Müller

STRUCTURE
The invitation to the one-day workshop read:
“the EUCog network (http://www.eucognition.org/) is organising a small one-day workshop on July 8th at Munich airport, dedicated to the image of cognitive systems research and to the communication of our 'story' to the public and the related communities, esp. in academia and industry. This important meeting will have significant impact on the image of cognitive systems research and the funding agenda in the near future - and we trust it will be intellectually interesting as well.”

We had 25 invited prominent participants (see below)

We met in plenary and heard 3 introductory statements:
  • The EC Unit “Robotics” and Cognitive Systems: Olivier da Costa 
  • The Robotics PPP and Cognitive Systems: Giorgio Metta
  • "Lessons and Ideas from Walking with Robots": Alan Winfield 
  • "EUCog and its message": Sebastián Grinschpun 

We then split into three working groups. The results of these three groups and the further actions are:

RESULTS

Group a): Lobbying
We need:
  • information gathering exercise from EUCog members (for ‘success stories’)
  • usable result map from relevant FP6/7 projects and the EUCog network
  • plan on how to continue EUCog network
  • contributions to the relevant EU roadmaps
We will:
  • propose a topic group for the PPP
  • provide input to the robotics SRA (strategic research agenda), MAR (multi-annual roadmap)
  • provide input for other funding initiatives (European or national)
    • collection of facts and ‘stories’
    • write white paper

Group b): Public Relations
Media
  • More use of social media (esp. YouTube, Facebook, Twitter)
  • More ‘open science’, blogging about activities
  • Better material (photos, videos, software to try out, interviews)
  • A focal person, a ‘celebrity scientist’ helps
  • Industry days for specific constituencies
Issues
  • Things that think, and provide insights into ourselves
  • Examples: autonomous driving, video games, voice recognition, intelligence in my pocket
  • Small key messages that are repeated (e.g. perception, autonomy, flexibility, learning, not pre-programmed, real-life)
  • Threats to be addressed (e.g. military use, job loss, ...)

Group c): Messages to Academia and Industry
Examples of situations that today’s robotics can’t handle, but cognitive systems would help solve
  • Fukushima (including ‘simple’ things like stairs, doors, valves)
  • grasp any time of any object (e.g. in a supermarket)
  • instruction ‘clean up’ doesn’t work
Examples of specific insights
  • Behavior is emergent, control + morphology
  • Grasping does not require knowing the shape
  • Comparison is more powerful than precision
  • The power of optic flow (rather than measuring distance and speed)
Examples of working systems
  • cameras with space-variant vision
  • robotic vacuum cleaners
  • Amazon user profiles
  • iPhone (sensing, concurrent, slam, ...)
  • offloading cognition to the environment
  • multimodal autonomous and assisted driving

Some Material Relevant for Future Work

Participants
Tjeerd Andringa
Tony Belpaeme
Joanna Bryson
Gordon Cheng
Olivier Da Costa
Andreas Engel
Christian Freksa
Toni Gomila
Sebastián Grinschpun
Matej Hoffmann
Lilla Lukacs
Colette Maloney
Giorgio Metta
Vincent C. Müller
Rolf Pfeifer
Florian Röhrbein
Yulia Sandamirskaya
Aaron Sloman
Hans-Georg Stork
David Vernon
Paul Verschure
Markus Vincze
Barbara Webb
Alan Winfield
Tom Ziemke