Sensory-motor representations and error learning - experimental analysis of manual intelligence in first order reality, virtual reality and augmented reality
Term: 2008-11 till 2012-08
Research Area: B A D
Abstract
One central issue for the cognitive control of movement is the compensation of errors and learning processes that enhance error compensation mechanisms. This is especially true for very precise movements such as many manual actions. The present project combines methods and conventional experimental settings (first order reality) with approaches from Virtual Reality and Augmented Reality to embed participants in interaction loops in which the occurrence and perception of errors can be manipulated and studied in novel ways.
Research Questions and Methods
To gain skilled behavior learners have to pass through stages of conscious, subconscious and unconscious internal processing of task related requirements. Thereby skill acquisition processes are characterized by rapid adaptations but with large movement errors in the initial stage which are reduced with an increasing amount of practice.
In the present project we aim to uncover sources of information that enhance motor learning processes to investigate what information is or might be integrated in sensory-motor representation to efficiently reduce movement errors. In most of the existing (ecological) sensory-motor learning settings learners are only able to use feedback information captured with their egocentric view. By the implementation of augmented or virtual reality where participants either receive task relevant information about movement kinematics additional to their egocentric view or only see their movement execution from an side-view angle, the present project wants to detect information that is sufficiently integrated in sensory-motor representation to support motor learning. It is likely that the reduction of movement errors can be expedited, (1) as well by providing additional information about movement kinematics online since this might intensify the learners’ internal link between perception and cognition (i.e. sensory-motor representation) and (2) by changing learners’ perspective from a goal-oriented view to an process-oriented view since that also provides the learner additional information that can be integrated in the sensory-motor representation more directly. A change in perspective and the visualization of movement kinematics might also improve erroneous behavior at late unconscious stages (e.g. anticipatory motor planning).
The project combines methods and conventional experimental settings (first order reality) with approaches from Virtual Reality and Augmented Reality (second order) to embed subjects in interaction loops in which the occurrence and perception of errors can be manipulated and studied in novel ways.
In motor learning settings for goal-directed action (e.g., precise aiming at far targets), which especially in the beginning result in performance errors, supporting information of the movement execution are introduced via Augmented Reality or Virtual Reality. Using Head mounted displays (HMD) participants are either only informed about their actual movement from a side-view angle (closed HMD) or about relevant kinematics of the ongoing trial additional to their normal egocentric view of the target (see-through HMD). All information are provided online during movement execution. A comparison of these different information conditions is expected to provide insights in error learning mechanisms. It might show, which information is required to efficiently reduce performance errors and update the sensory-motor representation.
Similarly, unconscious erroneous behavior in manual actions will be investigated with two different HMDs.
In this way we hope to gain new insights about error correction mechanisms, error compensation learning and their replication in technical systems such as robots.
« close
read more »
Outcomes
The project is designed to get detailed insights into error correction processes and error compensation learning. Using different types of Head mounted displays the learner gets online feedback from different sources either of his own movement from a side-view angle (closed HMD) or he is provided with additional kinematic data (i.e. actual vs. target values) during his normal egocentric view on the target (see-through HMD). The comparison of these two groups with a control group that learns the given manual task in a usual manner without supporting information, is expected to provide insights in the mechanisms of error corrections, i.e. which information is integrated in the sensory-motor representation (as supporting information to update the representation) for efficiently compensating movement errors.
First studies on that topic clearly show an advantage of learners that are additionally provided with relevant kinematics compared to a group of normal learners. Most interesting the reduction of performance error in a targeted throwing task was similar between the groups with see-through and closed HMD.
