Fault Diagnosis and Fault Tolerance for Mechanotronic Systems: Recent Advances

Fault Diagnosis and Fault Tolerance for Mechanotronic Systems: Recent Advances

Fault Diagnosis and Fault Tolerance for Mechanotronic Systems: Recent Advances

Fabrizio Caccavale and Luigi Villani (Eds)SpringerBerlin2003ISBN 3-540-44159-Xxii + 191 pp.hardback£49.00

Keywords: Fault tolerance, Information technology, Information systems

In a foreword to this first volume in a new series, the series editor indicates the purpose as being to disseminate information on new developments rapidly and informally but with high quality, as soon as a topic has reached the stage of maturity where such presentation is appropriate. Since this series is concerned with robotics he takes satisfaction in noting that this first volume has wide-ranging applicability including underwater vehicles and aircraft.

The term “mechanotronics” denotes the integration of mechanical, electronic and information technologies. In many applications, the consequences of faults are likely to be serious and automatic fault diagnosis is valuable. It is particularly valuable if the location and nature of the fault are indicated, and even more so if operation can be maintained despite it.

Application areas in which automatic fault diagnosis is specially valuable include unmanned vehicles for underwater exploration and operation in conditions that are dangerous for humans because of radioactivity or other dangerous conditions, as well as in space craft and planetary rovers. The breakdown of a robot in an area inaccessible to humans can obviously be very costly, and similar considerations apply to military aircraft, ships and vehicles. Prompt fault diagnosis and remediation are also important in industrial plants and it is mentioned that there is a plan to have motor vehicles in North America fitted with a form of automatic fault diagnosis to protect the environment. The authors of one of the papers visualise this being extended to boats and lawn mowers.

The six chapters of the book originated as presentations at a workshop on Fault Diagnosis and Fault Tolerance for Dynamic Systems, held in Vancouver in October 2002 in conjunction with an IEEE Symposium, and have been revised and expanded by the authors. All of them are fairly mathematical and acquaintance with modern control theory is assumed.

The first chapter is on “Sliding Mode Observers and Their Application in Fault Diagnosis”, where the sliding mode observers (SMOs) indicates a previously-existing body of theory that is introduced here extremely sketchily. A useful reference seems to be a book by Utkin, about which notes can be found at: http://www.eleceng.ohio-state.edu/,utkin/ee894v.html

The second chapter is also strongly theoretical and refers to “Fault Diagnosis and Fault Tolerant Control for Non-Gaussian Stochastic Systems with Random Parameters”.

The four remaining chapters refer to specific areas of application. One is on “Fault Diagnosis for Industrial Robots”, and another is a “Survey of Fault Detection/Tolerance for AUVs and ROVs” where the acronyms stand for autonomous underwater vehicle and remotely operated vehicle. In the chapter, the reference to ROVs is also to underwater operation. The discussion is highly practical with a listing of specific failures that might occur and pictures of an AUV and ROV and references to a successful AUV cable-laying mission in the Arctic that was only achieved because of fault-tolerant operation. The use of neural networks is mentioned in connection with this type of application.

The two remaining chapters are on “Failure Detection, Identification and Reconfiguration in Flight Control” and “Nonlinear Fault Detection for Hydraulic Systems”. The latter may seem somewhat specialized, but is important because hydraulic actuators are critical components of robotic devices that enter hazardous environments. The chapter on flight control is largely with reference to a tailless fighter aircraft in which failure tolerance is carried to the extreme where even physical damage to flight surfaces can be allowed for. This is only possible with redundancy of control, such that the available control actions have extra degrees of freedom. The switch to a new control mode is made quickly and automatically following damage.

The book delineates what is obviously an extremely important development area. A surprising aspect, especially in view of the mention of information technology, is that there seems to be no reference whatsoever to the well explored area of error detecting and error correcting coding in digital communication, nor to the extension of the principles to reliable neural computing as first suggested by McCulloch (1959) and later developed by his associates. It would be surprising if there was not some basis for cross-fertilization here.

Alex M. Andrew

ReferenceMcCulloch, W.S. (1959), “Agatha Tyche – of nervous nets, the lucky reckoners”, Mechanisation of Thought Processes, Proceedings of a Symposium in the National Physical Laboratory, Teddington, HMSO, London.