Keywords
Citation
Harwood, C.J. (1999), "Lecture Notes in Artificial Intelligence 1416. Subseries of Lecture Notes in Computer Science. Tasks and Methods in Applied Artificial Intelligence", Kybernetes, Vol. 28 No. 9, pp. 1084-1094. https://doi.org/10.1108/k.1999.28.9.1084.3
Publisher
:Emerald Group Publishing Limited
The second volume of the proceedings of the 11th International Conference on Industrial and Engineering Applications of Artificial Intelligence and Expert Systems – IEA‐98‐AIE, held at Benicàssim, Castellón, Spain, June 1‐4, 1998 has the title: Tasks and Methods in Applied Artificial Intelligence. Volume I; Methodology and Tools in Knowledge‐Based Systems has already been reviewed in this section.
The division of the conference proceedings into these two sections was cleanly done and Volume II concentrates on the contributions dealing with aspects that are more directly relevant to applications development. The best insight into the types of applications is seen by examining the grouping of the contributions into the five parts and their subtitled topics. Contents were arranged as:
- 1.
(1) Synthesis tasks (two session papers) with the sub‐topics:
- •
Spatial, temporal and spatio‐temporal planning and scheduling (four papers)
- •
Motion planning for robots (eight papers)
- •
System configuration (one paper)
- 1.
(2) Modification tasks – with the sub‐topics:
- •
Knowledge‐based control systems (five papers)
- •
Dynamic systems supervision (five papers)
- •
Intelligent supervision (two papers)
- •
Predictive control systems (one paper)
- •
Perceptual robotics (nine papers)
- •
Fuzzy and neurofuzzy approaches to robot control (two papers)
- •
Program reuse (four papers)
- 1.
(3) Machine learning:
- •
Machine learning applications: tools and methods (five papers)
- •
Inductive and deductive strategies (one paper)
- •
Case‐based reasoning (six papers)
- •
Learning advances in neural networks (seven papers)
- 1.
(4) Applied Artificial Intelligence and knowledge‐based systems in specific domains (1 session paper):
- •
Multimedia (two papers)
- •
Human‐computer interaction (three papers)
- •
Decision support systems (four papers)
- •
Tutoring systems (three papers)
- •
Connectionist and hybrid AI approaches to manufacturing (four papers)
- •
Modeling and simulation of ecological/environmental systems (five papers)
- •
Reports on real applications with significant findings (four papers)
- 1.
(5) Validation and evaluation criteria (three session papers):
- •
Simulation (four papers)
These titles give an insight to the range of topics presented and indeed to the popularity of some of the subjects with participants and, once again, with the selectors and referees chosen by the conference organisers. At least the organisers were aware, as the editors of Volume II were, that there are grand challenges for AI concerning artificial behaviour for agents that have to deal with the real world through perception and motor actions. Most users of systems know of the great lack of balance between existing AI endeavours and in some aspects of their competence. The editors of this text make this point in their preface to both volumes. They neatly sum this up by writing:
Whereas in some formal microworlds AI systems have reached the highest human level of competence – the recent success of chess playing systems being a paradigmatic example, or there are knowledge‐based systems exhibiting human expert competence in narrow technical domains such as medical diagnostics etc., few systems exist surpassing the competence of the cockroach, for instance in moving around pursuing a goal in an unstructured world.
This is why most scientists see the great gap between the pure abstract intellectual tasks, on the one hand, and those that involve sensor motor interaction with the physical world, on the other. The editors of this volume, quite rightly, call for an emphasis on research on robotic agents. They blame the Turing Vision, that is the vision of a more disembodied, abstract, symbol‐processing intelligence. Robotic capacities have been extended to the Turing test and new approaches made. Indeed studies in robotics have not been neglected in this text and the section in group 1 on Motion Planning for Robots could be the basis for a conference in its own right. Similarly, in the Group 3 the section on Perceptual Robotics has a wealth of detail that could well have been enlarged upon but for the obvious constraints of both conference time and space in the published proceedings. The volume also has mention of the fuzzy and neurofuzzy approaches to robot control and it was disappointing to see only two papers on this essential topic. The papers published on Neural Networks (seven papers) could only be described as providing awareness of an approach to problems in a subsymbolic manner. Its strength will be as one perspective among others that should be part of a mutually supporting strategy for problem solution.
It is interesting to note the editors’ comments that the emphasis on perception and robotics had obtained a satisfactory response in terms of the number of submitted papers, compared with previous conferences. Not all of those published in this volume would have been selected by hard‐nosed journal editors in the robotics field for inclusion in their publications. Even so, this volume does follow the conference theme enthusiastically, and new methodologies, knowledge modelling and hybrid techniques were well focused and are truly representative of current activity in the fields.
The editors of the text are right in believing that the global assessment of these published contributions will be positive and it should certainly be recommended to researchers and developers in these fields. Since it is a representative “state‐of‐the‐art” volume its shelf life will be limited even so it does make a major contribution to Applied Artificial Intelligence and Knowledge Engineering and is worth consulting both for reference and for an overview of what is still a new and fascinating study area.