Aleksandar Vakanski

TRAJECTORY LEARNING BASED ON CONDITIONAL RANDOM FIELDS FOR ROBOT PROGRAMMING BY DEMONSTRATION

This work presents an approach for implementation of conditional random fields (CRF) in transferring motor skills to robots. As a discriminative probabilistic model, CRF models directly the conditional probability distribution over label sequences for given observation sequences. Hereby, CRF was employed for segmentation and labeling of a set of demonstrated trajectories observed by a tracking sensor. The key points obtained by CRF segmentation of the demonstrations were used for generating a generalized trajectory for the task reproduction. The approach was evaluated by simulations of two industrial manufacturing applications.

A secure architecture for IoT with supply chain risk management

We proposes the development of a cyber-secure, Internet of Things (IoT), supply chain risk management architecture. The proposed architecture is designed to reduce vulnerabilities of malicious supply chain risks by applying machine learning (ML), cryptographic hardware monitoring (CHM), and distributed system coordination (DSC) techniques to mitigate the consequences of unforeseen (including general component failure) threats. In combination, these crosscutting technologies will be integrated into Instrumentation-and-Control/Operator-in-the-Loop (ICOL) architecture to learn normal and abnormal system behaviors. The detection of absolute or perceived abnormal system-component behaviors will trigger an ICOL alert that will require an operators manual verification-response action (i.e., that the detected alert is, or is not a viable control system threat). The operators verification-response will be fed back into the ML systems to recalibrate the perceived normal or abnormal state of the systems behavior.

Robotic Learning of Manipulation Tasks from Visual Perception Using a Kinect Sensor

—The paper addresses the problem of transferring new skills to robots from observation of human demonstrated skill examples. An approach is presented for retrieving trajectories of an object, being manipulated during the demonstrations, from Kinect-provided measurements. The problem of object tracking across the image frames is solved by using weighted dynamic template matching with normalized cross-correlation. Such approach takes advantage of the simultaneous image and depth measurements by the Kinect device in leveraging the pattern localization and pose estimation. Demonstrated trajectories are stochastically encoded with hidden Markov model, and the obtained model is exploited for generation of a generalized trajectory for task reproduction. The developed methodology is experimentally validated in a real-world task learning scenario.

A Data Set of Human Body Movements for Physical Rehabilitation Exercises

The article presents University of Idaho – Physical Rehabilitation Movement Data (UI-PRMD) — a publically available data set of movements related to common exercises performed by patients in physical rehabilitation programs. For the data collection, 10 healthy subjects performed 10 repetitions of different physical therapy movements, with a Vicon optical tracker and a Microsoft Kinect sensor used for the motion capturing. The data are in a format that includes positions and angles of full-body joints. The objective of the data set is to provide a basis for mathematical modeling of therapy movements, as well as for establishing performance metrics for evaluation of patient consistency in executing the prescribed rehabilitation exercises.

An image-based trajectory planning approach for robust robot programming by demonstration

Abstract The article proposes a new robot programming-by-demonstration framework, which integrates a visual servoing tracking control to robustly follow a trajectory generated from observed demonstrations. The constraints originating from the use of a visual servoing controller are incorporated into the trajectory learning phase, to guarantee feasibility of the generated plan for task execution. The observational learning is solved as a constrained optimization problem, with an objective to generalize from a set of trajectories of salient features in the image space of a vision camera. The proposed approach is evaluated experimentally for learning trajectories acquired from kinesthetic demonstrations.