Your personal computer experts

Who we are...

Help! Equip your office Automate your office Develop your web presence

   Who we are

   Our references

   Price list


If you would like to meet with us personally to discuss your needs, please give us a call!


Kraan and Richards
Entfelderstrasse 7
5012 Schönenwerd

Tel 062/849 9746
Fax 062/849 9743


Bradley L. Richards: Curriculum Vitae

Education

High School (August 1974 through May 1978)
     Albuquerque Academy, Albuquerque, New Mexico, USA
     Cum Laude, Who's Who 1977-78

B.Sc. in Electrical Engineering (August 1978 through May 1982)
     Southern Methodist University, Dept. of Electrical Engineering, Dallas, Texas, USA

United States Air Force, Officer Training School (June 1982 through August 1982)
     San Antonio, Texas, USA

M.Sc. in Electrical Engineering (May 1985 through December 1986)
     Air Force Institute of Technology, Dept. of Electrical Engineering, Dayton, Ohio, USA
     Thesis: Programming in Fuzzy Logic: Fuzzy Prolog
     Louis F. Polk award, Distinguished Graduate
     Who's Who 1985-86 and 1986-87
     Vice-president, Student Council

United States Air Force, Squadron Officer's School (April 1988 through May 1988)
     Montgomery, Alabama, USA

Ph.D. and M.Sc. in Computer Science (September 1989 through August 1992)
     University of Texas, Dept. of Computer Sciences, Austin, Texas, USA
     Dissertation: An Operator-Based Approach to First-Order Theory Revision
     Supervisor: Professor Raymond J. Mooney
     Board Member, Graduate Student Council in Computer Science

Professional Experience

Johnson Controls, Inc. (Dallas, Texas, USA; January 1979 through May 1982). Cooperative education program with industry. Wrote real-time process-control software for sewage plant automation. Responsible for software documentation and project configuration control.

United States Air Force, F-15 System Program Office (Dayton, Ohio, USA; September 1982 through May 1985). Software Systems Manager. Reviewed technical specifications, negotiated costs, and accepted final delivery of $50 million of avionics and ground support software for the F-15 fighter aircraft. Established and reviewed contractor and subcontractor software quality assurance programs.

Defense Logistics Agency (Boston, Massachusetts, USA; January 1987 through August 1989). Program Support Officer. On-site Air Force representative to Raytheon Corporation's 27 plants in the Boston area, for $2 billion of Air Force contracts. Simultaneously served as project engineer for the Advanced Medium Range Air-to-Air Missile (AMRAAM) Producibility Enhancement Program.

Merrimack College (North Andover, Massachusetts, USA; January 1987 through August 1989). Part-time Lecturer. Taught courses in Pascal and Compilers. Chaired the Computer and Information Sciences Curriculum Committee, which developed a new computer science curriculum to meet ACM accreditation standards.

University of Edinburgh, Dept. of Artificial Intelligence (Edinburgh, Scotland; October 1992 through December 1992). Research contract to implement the Mollusc proof development system. Mollusc provides logic-independent support for interactive and semi-automatic theorem proving, and serves as an object-level verification system for proof planning and program synthesis systems.

University of Aberdeen, Dept. of Computing Sciences (Aberdeen, Scotland; January 1993 through August 1993). Research Fellow on the VIVA (Verification, Improvement, and Validation of Knowledge-Based Systems) Esprit project. Development of translation tools to convert between commercial expert-system shell representations and the common knowledge representation used by VIVA. Extension of techniques developed in experimental knowledge refinement systems such as KRUST and FORTE to work with industrial expert systems.

Swiss Federal Institute of Technology (EPFL), Artificial Intelligence Laboratory (Lausanne, Switzerland; November 1993 through through March 1995). Research Fellow in the Working Group on Model-Based Reasoning (Schwerpunkt program). Development of a model-based indexing method which allows case-based reasoning to be applied to complex process-control problems. Proof of concept through creation and evaluation of a prototype tool to automatically develop qualitative models from process data, and to derive case indices from the resulting model.

University of Applied Sciences in Furtwangen (FHF), Department of Computer Science (Furtwangen, Germany; April 1995 through February 2001). Professor for software engineering and artificial intelligence, with lectures in these and related areas. In 1996, together with Prof. Mescheder (also at the Fachhochschule Furtwangen), leader of a research project to develop an expert system for microsystem design; this project was funded by the Baden-Württemberg Ministry for Science and Research. I was a member of the Curriculum Committee, chairman of two search committees, and from 1998 through 2001 deputy head of department.

Engineering School of Basel (FHBB), Department of Computer Science (Basel, Switzerland; February 2001 through present). Part-time Lecturer in Computer Science. I give lectures in Theoretical Foundations of Computer Science, Internet Programming, and Java. In addition, I supervise student projects.

Kraan and Richards (Schönenwerd, Switzerland; January 1994 to present). Co-owner. This company is active in two very different fields. First, the import and retailing of Scottish products, particularly Scotch whisky, into Switzerland. Second, we consult for small companies that want to make better use of computer technology, we develop custom back-office applications and develop their web presence on the Internet. Our principle product in this area is the KVS software for school administration.

Five Most Important Research Publications

I have not been active in research since 1996

B. Richards, B. Faltings, and P. Duxbury-Smith, “Case-based Modeling with Qualitative Indices”, Proceedings of the International Joint Conference on Artificial Intelligence (IJCAI-95), 1995. We build a partial or complete qualitative model of a physical system, and use this model to derive the minimal set of parameters relevant to each system input. This reduces the number of attributes used for indexing, thereby increasing the density of cases in the parameter space and allowing effective predication and control of the system.

B. L. Richards and R. J. Mooney, "Refinement of First-Order Horn-Clause Domain Theories," Machine Learning Journal, 19:2, 1995). Knowledge acquisition is a difficult and error-prone task. This paper presents an implemented system that automatically detects and localizes errors in a knowledge base, based on examples of desired behavior. It then uses a library of transformation operators to correct these errors, producing a theory consistent with the given examples. The system is tested and evaluated in a wide variety of domains.

B. L. Richards, I. Kraan, A. Smaill, and G. A. Wiggins, "Mollusc: A General Proof-Development Shell for Sequent-Based Logics," Proceedings of the 1994 Conference on Automated Deduction, (CADE-94), 1994. Mollusc is a proof development system which can be used to construct and edit proofs in sequent-based logics. In addition to proof editing facilities, Mollusc supports the definition of new logics, includes a proof-planner interface, and provides for automated proof construction through a tactic language and interpreter.

B. L. Richards and R. J. Mooney, "Learning Relations by Pathfinding," Proceedings of the Tenth National Conference on Artificial Intelligence (AAAI-92), 1992. First-order machine learning systems generally use a form of hill-climbing to avoid combinatorial explosion when developing possibly changes to a theory. This article presents a highly effective heuristic that helps systems avoid local maxima and local plateaus in the search space.

B. L. Richards, I. Kraan, and B. J. Kuipers, "Automatic Abduction of Qualitative Models," Proceedings of the Tenth National Conference on Artificial Intelligence (AAAI-92), 1992. We describe an implemented method that can automatically derive a provably-correct qualitative system model from qualitative or quantitative descriptions of system behavior.