NIFTi: Natural human-robot cooperation in dynamic environments
NIFTi is about human-robot cooperation. About teams of robots and humans acting together, interacting together, performing tasks together, trying to reach a shared goal.
Much research has gone into how robots could function autonomously, as part of such teams. Work on their own, moving around, executing their assignments.
NIFTi complements this autonomy in operation, with a perspective on human-robot teaming. Little has been said so far on how a robot could cooperate with a human. Not just having a human in the loop -- but actually bearing the human in mind, when determining what to do or say next in human-robot team interaction; when, and how.
This is where NIFTi comes in. NIFTi puts the human factor into cognitive robots, and into human-robot (team) interaction in particular.
NIFTi focuses on tasks in Urban Search & Rescue. Human-robot teams work together to explore a disaster area, to assess the situation, locate victims.
In this context NIFTi investigates how a cognitive robot can complement models of its own capabilities and situation awareness, with cognitive user models of task load and workflow. NIFTi works off the hypothesis that a combination of these can achieve natural human-robot cooperation. The idea is to continuously balance how to operate autonomously, with how to interact with others in the team so as to minimize task load for human team members, and optimize workflow. NIFTi calls this the naturalness loop.
To bring this about, NIFTi does not see user models as just another box in an architecture scheme. On the contrary. The human factor pervades most of the robot’s models and decision procedures. How it understands the environment. How it combines its situation awareness with user models to estimate what a human might pay attention. How a situation may affect a human’s task load. To estimate, basically, how a human may act or react in a dynamic environment. So that the robot can accordingly try to adapt what it does, decides to do or say next in interacting with humans, when, and how.
So that the robot can bear the human in mind when acting. Literally.
NIFTi will result in an integrated account of how natural human-robot teaming can come about. This account will be instantiated in several generations of integrated systems. Each year, systems will be evaluated with several USAR end user organizations. Rescue personal will team up with the NIFTi robots to carry out realistic missions, in real-life training areas. NIFTi will develop a novel rover platform to meet the demands of operating in such environments.
NIFTi is an Integrated Project funded by the EU FP7 ICT Programme, in the Cognitive Systems & Robotics unit (Project #247870). NIFTi runs from January 1 2010 until December 31 2013.
NIFTi is coordinated by DFKI GmbH (Geert-Jan M. Kruijff). NIFTi comprises DFKI, TNO Human Factors, Fraunhofer IAIS, BlueBotics, ETH Zürich, Czech Technical University, “La Sapienza” University of Rome as scientific partners. NIFTi proudly cooperates with the Fire Department of Dortmund, the Italian national fire and rescue services, the rescue command of the Swiss Army, and RUAG, as end user organizations.