4 Matching Results

Search Results

Advanced search parameters have been applied.

Artificial awareness for robots using artificial neural nets to monitor robotic workcells

Description: Current robotic systems are unable to recognize most unexpected changes in the work environment, such as tool breakage, workpiece motion, or sensor failure. Unless halted by a human operator, they are likely to continue actions that are at best inappropriate, and at worst may cause damage to the workpiece or robot. This project investigated use of Artificial Neural Networks (ANNs) to learn the expected characteristics of sensor data during normal operations, recognize when data no longer is consistent with normal operation, suspend operations and alert a human operator. Data on force and torque applied at the robot tool tip were collected from two workcells: a robotic deburring system and a robot material-handling system. Data were collected for normal operations and for operations in which a fault condition was introduced. Data simulating sensor failure and excessive sensor noise were generated. Artificial Neural Networks (ANN) were trained to classify operating conditions; several ANN architectures were tested. The selected ANNs were able to correctly classify all valid operating conditions and the majority of fault conditions over the entire range of operating conditions, having {open_quotes}learned{close_quotes} the expected force/torque data. Most faults introduced appreciable error in the data; these were correctly classified. However, a small minority of faults did not give rise to a detectable difference in force and torque data. It is believed that these faults could be detected using other sensors. The computational workload varies with the implementation, but is moderate: up to 2.3 megaflops. This makes implementation of a real-time workcell monitor feasible.
Date: April 1, 1997
Creator: Tucker, S.D. & Ray, L.P.
Partner: UNT Libraries Government Documents Department

Knowledge assistant: A sensor fusion framework for robotic environmental characterization

Description: A prototype sensor fusion framework called the {open_quotes}Knowledge Assistant{close_quotes} has been developed and tested on a gantry robot at Sandia National Laboratories. This Knowledge Assistant guides the robot operator during the planning, execution, and post analysis stages of the characterization process. During the planning stage, the Knowledge Assistant suggests robot paths and speeds based on knowledge of sensors available and their physical characteristics. During execution, the Knowledge Assistant coordinates the collection of data through a data acquisition {open_quotes}specialist.{close_quotes} During execution and post analysis, the Knowledge Assistant sends raw data to other {open_quotes}specialists,{close_quotes} which include statistical pattern recognition software, a neural network, and model-based search software. After the specialists return their results, the Knowledge Assistant consolidates the information and returns a report to the robot control system where the sensed objects and their attributes (e.g. estimated dimensions, weight, material composition, etc.) are displayed in the world model. This paper highlights the major components of this system.
Date: December 1, 1996
Creator: Feddema, J.T.; Rivera, J.J. & Tucker, S.D.
Partner: UNT Libraries Government Documents Department

Optical and control modeling for adaptive beam-combining experiments

Description: The development of modeling algorithms for adaptive optics systems is important for evaluating both performance and design parameters prior to system construction. Two of the most critical subsystems to be modeled are the binary optic design and the adaptive control system. Since these two are intimately related, it is beneficial to model them simultaneously. Optic modeling techniques have some significant limitations. Diffraction effects directly limit the utility of geometrical ray-tracing models, and transform techniques such as the fast fourier transform can be both cumbersome and memory intensive. The authors have developed a hybrid system incorporating elements of both ray-tracing and fourier transform techniques. In this paper they present an analytical model of wavefront propagation through a binary optic lens system developed and implemented at Sandia. This model is unique in that it solves the transfer function for each portion of a diffractive optic analytically. The overall performance is obtained by a linear superposition of each result. The model has been successfully used in the design of a wide range of binary optics, including an adaptive optic for a beam combining system consisting of an array of rectangular mirrors, each controllable in tip/tilt and piston. Wavefront sensing and the control models for a beam combining system have been integrated and used to predict overall systems performance. Applicability of the model for design purposes is demonstrated with several lens designs through a comparison of model predictions with actual adaptive optics results.
Date: August 1, 1995
Creator: Gruetzner, J.K.; Tucker, S.D.; Neal, D.R.; Bentley, A.E. & Simmons-Potter, K.
Partner: UNT Libraries Government Documents Department

Multi-segment coherent beam combining

Description: Scaling laser systems to large sizes for power beaming and other applications can sometimes be simplified by combing a number of smaller lasers. However, to fully utilize this scaling, coherent beam combination is necessary. This requires measuring and controlling each beam`s pointing and phase relative to adjacent beams using an adaptive optical system. We have built a sub-scale brass-board to evaluate various methods for beam-combining. It includes a segmented adaptive optic and several different specialized wavefront sensors that are fabricated using diffractive optics methods. We have evaluated a number of different phasing algorithms, including hierarchical and matrix methods, and have demonstrated phasing of several elements. The system is currently extended to a large number of segments to evaluate various scaling methodologies.
Date: December 31, 1994
Creator: Neal, D.R.; Tucker, S.D.; Morgan, R.; Smith, T.G.; Warren, M.E.; Gruetzner, J.K. et al.
Partner: UNT Libraries Government Documents Department