Grant: Discovery Grants Program - Individual
University of Manitoba
$14,000.00
Active balancing and complex motion planning for adult sized humanoid robots
EF 2012-13
Résumé
The generation of stable walking and motion planning for adult sized (> 1.4m) humanoid robots presents unique challenges, because the state space is high-dimensional, the available sensor feedback is limited, and the robots are underactuated. This proposal aims at investigating new approaches to fundamental problems involving these robots. First, how can we design and implement algorithms so that humanoid robots cannot only move over flat even surfaces, but also move over uneven terrain? The solution to this problem will allow robots to travel across grass surfaces, forest paths, and rugged terrain. Second, how can we develop algorithms so that robots can plan novel motions from scratch rather than having to pre-program them? This will allow humanoid robots to pick up objects, plan its footsteps, and to even climb up a rock wall.
Both areas have received significant attention in recent years, and progress has been made - but versatile, robust, practical solutions are still elusive. The research described in this proposal aims at greatly increasing the range of balancing and actions that a robot can execute by allowing some of its movements to be open-loop rather than closed-loop.
To overcome the need for highly accurate global models of the kinematics and dynamics of the robot as well as its interaction with the environment, we propose to use multiple models with different resolutions that model specific important parts of the behaviour of the robot.
Applying this idea to active balancing will allow us to overcome restrictions imposed by the 3D inverted pendulum model and the zero moment point (ZMP) constraint and to generate more stable, faster, more efficient and more natural looking walking gaits and balancing behaviours. The controller will generate balancing reflexes that will bring an unbalanced robot back into the controllable balanced region.
We propose novel motion planning algorithms that uses the controllable balanced regions rather than regions based on kinematic and dynamic constraints as state space for the planning process.
Programme
Discovery Grants Program - Individual
Bénéficiaire
Emplacement
Manitoba, Canada
Type de bénéficiaire
Academia
Type
Grant
Source
NSERC's Awards Data
Tel que publié
Application id
262062-2011
Name nom
Baltes, Hansjorg(Jacky)
Department département
Computer Science
Organization id
19
Institution établissement
University of Manitoba
Province en
Manitoba
Province fr
Manitoba
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2012
Competition year année de concours
2011
Award amount
14000
Program id
RGPIN
Program name en
Discovery Grants Program - Individual
Program name fr
Programme de subventions à la découverte - individuelles
Group en
Discovery grants programs
Group fr
Programmes de subventions à la découverte
Committee code
1507
Committee name en
Computer Science
Committee name fr
Sciences informatiques
Area of application code
1100
Area of application group en
Health, education and social services
Area of application group fr
Santé publique, éducation et services sociaux
Area of application en
Health, education and social services
Area of application fr
Santé publique, éducation et services sociaux
Research subject code
2600
Research subject group en
Robotics
Research subject group fr
Robotique
Research subject en
Robotics
Research subject fr
Robotique
Partie
2
Nb partie
5
Application title
Active balancing and complex motion planning for adult sized humanoid robots
Keyword
NSERC
Application summary
The generation of stable walking and motion planning for adult sized (> 1.4m) humanoid robots presents unique challenges, because the state space is high-dimensional, the available sensor feedback is limited, and the robots are underactuated. This proposal aims at investigating new approaches to fundamental problems involving these robots. First, how can we design and implement algorithms so that humanoid robots cannot only move over flat even surfaces, but also move over uneven terrain? The solution to this problem will allow robots to travel across grass surfaces, forest paths, and rugged terrain. Second, how can we develop algorithms so that robots can plan novel motions from scratch rather than having to pre-program them? This will allow humanoid robots to pick up objects, plan its footsteps, and to even climb up a rock wall.
Both areas have received significant attention in recent years, and progress has been made - but versatile, robust, practical solutions are still elusive. The research described in this proposal aims at greatly increasing the range of balancing and actions that a robot can execute by allowing some of its movements to be open-loop rather than closed-loop.
To overcome the need for highly accurate global models of the kinematics and dynamics of the robot as well as its interaction with the environment, we propose to use multiple models with different resolutions that model specific important parts of the behaviour of the robot.
Applying this idea to active balancing will allow us to overcome restrictions imposed by the 3D inverted pendulum model and the zero moment point (ZMP) constraint and to generate more stable, faster, more efficient and more natural looking walking gaits and balancing behaviours. The controller will generate balancing reflexes that will bring an unbalanced robot back into the controllable balanced region.
We propose novel motion planning algorithms that uses the controllable balanced regions rather than regions based on kinematic and dynamic constraints as state space for the planning process.