Scalable robot fault detection and identification
Introduction
In this paper a fault is defined as a deviation from expected behavior. Experience has shown that even carefully designed and tested robots may encounter faults [6]. One of the reasons for this is that components degrade over time. Another is that the developers of the robot rarely have complete knowledge of the environment in which it operates and hence may not have accounted for certain situations.
Fault Detection and Identification (FDI) for robots is a complex problem. This is because the space of possible faults is very large, robot sensors, actuators, and environment models are uncertain, and there is limited computation time and power.
The algorithm presented in this paper uses Monte Carlo methods to gain accuracy. Classical Monte Carlo methods for dynamic systems, such as particle filters, are capable of tracking complex non-linear systems with noisy measurements. The problem is that estimates from a particle filter tend to have a high variance for small sample sets. Using large sample sets is computationally expensive and defeats the purpose.
This paper presents an approach for improving the accuracy of fault monitoring with a computationally tractable set of samples in a particle filter. The combination of two algorithms described in this paper enables monitoring of a wider range and larger number of faults during robot operation than has hitherto been possible. It can handle noisy sensors, non-linear, non-Gaussian models of behavior, and is computationally efficient.
Section snippets
Robot fault detection, identification, and monitoring
A fault is defined as a deviation from the expected behavior of the system. A failure is a complete interruption of the system’s ability to perform the required operation [12]. Fault detection is defined as the process of determining that a fault has occurred. Fault identification is the process of determining exactly which exception or fault occurred. Fault detection and identification are typically passive, i.e., they do not alter control actions. Fault monitoring is the process of providing
Classical particle filter for monitoring faults
Our formulation of the fault monitoring problem requires estimating the robot and environmental state, as it changes over time, from a sequence of sensor measurements that provide noisy, partial information about the state. The Bayesian approach to dynamic state estimation addresses this problem. Computing the exact Bayesian posterior analytically is intractable for the fault monitoring problem. Hence, we use a particle filter approximation in this paper. Particle filters are a Monte Carlo
Decision-theoretic particle filter
The decision-theoretic particle filter (DTPF) [22], [21] generates particles by factoring in cost to efficiently track rare high-risk events.
Faults are low-probability, high-cost events. The classical particle filter (CPF) generates particles proportional only to the posterior probability of an event. It is insensitive to costs that might arise from the particle approximation. Monitoring a system to detect and identify faults based on a CPF therefore requires a very large number of particles,
Decision-theoretic variable resolution particle filter
The variable resolution particle filter (VRPF) [23] introduces the notion of abstract particles, in which a particle may represent an individual state or sets of similar states. With this method, a single abstract particle simultaneously tracks multiple similar states. A limited number of particles are therefore sufficient for representing large portions of the state space when likelihood of occupying this part of the state space is low. When the likelihood of the grouped states increases and
Experimental setup
The experiments in this paper were performed using the Darwin2K simulator. Darwin2K is a free, open-source toolkit for robot simulation and automated design [15].
Darwin2K’s simulation capabilities are tailored to support engineering design and controller prototyping for robotic application. For example, it includes detailed motor and gear head models and provides full dynamic simulation capabilities.
A six-wheel rocker-bogie rover with actuated steering was used in the experiments. Fig. 1 shows
Acknowledgments
We thank Geoff Gordon and Sebastian Thrun for invaluable advice.
Reid Simmons is a Research Professor in the School of Computer at Carnegie Mellon University. He earned his B.A. degree in 1979 in Computer Science from SUNY at Buffalo, and his M.S. and Ph.D. degrees from MIT in 1983 and 1988, respectively, in the field of Artificial Intelligence. Since coming to Carnegie Mellon in 1988, his research has focused on developing self-reliant robots that can autonomously operate over extended periods of time in unknown, unstructured environments. This work
References (23)
A stochastic simulation bayesian approach to multitarget tracking
IEEE Transaction on Aerospace and Electronic systems
(1995)- et al.
Control of discrete-time hybrid stochastic systems
IEEE Transactions on Automatic Control
(1988) - et al.
Tracking and Data Association
(1988) - et al.
Dante ii: Technical description, results and lessons learned
International Journal of Robotics Research
(1999) - C. Bererton, P. Khosla, An analysis of cooperative repair capabilities in a team of robots, in: Proceedings of the IEEE...
- J. Carlson, R. Murphy, Reliability analysis of mobile robots, in: Proceedings of the IEEE International Conference on...
- A. Doucet, D. Crisan, A survey of convergence results on particle filtering for practitioners, in: IEEE Trans. Signal...
- et al.
Markov localization for mobile robots in dynamic environments
Journal of Artificial Intelligence Research
(1999) - et al.
Novel approach to nonlinear/non gaussian bayesian state estimation
IEE Proceedings-F
(1993)
CONDENSATION: conditional density propagation for visual tracking
International Journal of Computer Vision
Cited by (25)
A remote anomaly detection system for Slocum underwater gliders
2021, Ocean EngineeringCitation Excerpt :The pervasive adoption of Marine Autonomous Systems (MAS) is currently constrained by the challenges of operating fully independently of support vessels, in the demanding dynamic environment of the world’s oceans. MAS are piloted and monitored by experts, keeping operational costs high and limiting the scale of observational fleets that can be deployed simultaneously (Verma and Simmons, 2006). Since a MAS can be at sea for months at a time, operating ‘over-the-horizon’ from human pilots and support vessels, it is impossible to perform reactive maintenance, i.e. maintenance when a fault occurs.
Enhancing fault tolerance of autonomous mobile robots
2015, Robotics and Autonomous SystemsCitation Excerpt :This scientific topic is rarely addressed in major robotics conferences (ICRA, IROS). Concerning Robotics and Autonomous Systems, it seems that only one article has focused on the dependable issue [7] since 2006. There is also a need for a methodology to consciously guide robot designers to introduce dependable principles.
Fault detection on robot manipulators using artificial neural networks
2011, Robotics and Computer-Integrated ManufacturingCitation Excerpt :Verma and Simmons were presented a new algorithm for computationally tractable fault diagnosis. Experiments with a dynamic simulation of a six-wheel rocker-bogie rover showed a significant improvement in performance over the classical approach [9]. Fault tolerant motion planning of two spatial coordinating manipulators having two possible locked joints, respectively, was researched.
Fault Diagnosis Method for Industrial Robots Based on DBN Joint Information Fusion Technology
2022, Computational Intelligence and NeuroscienceFault Detection and Mutual Coordination in Various Cyborgs
2018, Lecture Notes in Networks and Systems
Reid Simmons is a Research Professor in the School of Computer at Carnegie Mellon University. He earned his B.A. degree in 1979 in Computer Science from SUNY at Buffalo, and his M.S. and Ph.D. degrees from MIT in 1983 and 1988, respectively, in the field of Artificial Intelligence. Since coming to Carnegie Mellon in 1988, his research has focused on developing self-reliant robots that can autonomously operate over extended periods of time in unknown, unstructured environments. This work involves issues of robot control architectures that combine deliberative and reactive control, probabilistic planning and reasoning, monitoring and fault detection, and robust indoor and outdoor navigation. More recently, Dr. Simmons has focused on the areas of coordination of multiple heterogeneous robots, human-robot social interaction, and formal verification of autonomous systems. Over the years, he has been involved in the development of over a dozen autonomous robots.
Vandi Verma is a Research Scientist at NASA Ames Research Center with QSS Inc. Her primary research interests are AI, robotics, and machine learning for space exploration. She has a Ph.D. in robotics from Carnegie Mellon University. This article was submitted while she was a graduate student at Carnegie Mellon University.