- Shopping Bag ( 0 items )
This book has been designed to provide professional engineers, scientists, and students with a consistent and practical framework for the rigorous conduct and communication of complex research and development projects. Although courses and training in research methods are common and generally required of social science professionals, a vast majority of physical scientists and engineers have had no formal classroom training or on-the-job mentoring on proper procedures for research methods. Getting It Right emphasizes the comprehensive analysis of project problems, requirements, and objectives; the use of standard and consistent terminology and procedures; the design of rigorous and reproducible experiments; the appropriate reduction and interpretation of project results; and the effective communication of project design, methods, results, and conclusions.
Presents a standard methodology for conducting coherent, rigorous, comprehensible, and consistent R&D projects
Thoroughly researched to appeal to the needs of R&D engineers and scientists in industry
Will also appeal to students of engineering and science
Audience: Scientists and engineers in all fields of research and development, social scientists, and managers of R&D projects.
At that time, Bosch Corporate Research was investigating pattern recognition and classification methods to process signals, images, and three-dimensional data representations like sonograms. Peter, who was on sabbatical leave from his home institution, The George Washington University in Washington DC, had been asked to set up a research project at the newly established Research Institute for Applied Knowledge Processing (FAW) in Ulm, Germany, not far from Schloss Reisensburg. The objective of the research project was to design and build a large-scale system for sophisticated image and signal processing, based on Collective Learning Systems Theory, a supervised adaptive learning paradigm that Peter had first proposed in the middle 1970s and has been developing and applying ever since.
The new image-processing engine that his research team was building at the FAW, called ALISA, was implemented on a Transputer-based parallel-processing computing engine. Each Transputer was programmed as a collective learning cell whose computational power could be compared to a few hundred neurons. In this way, a relatively powerful system could be achieved with existing hardware, overcoming the connectivity limits of a typical multilayer perceptron.
I was impressed by the rigor with which Peter had designed the ALISA engine and the project at the FAW. To implement his design, he brought together a group of graduate students from The George Washington University and scientists and engineers from Bosch. Strict adherence to the Scientific Method was the guiding methodological principle of the project. It was an enlightening experience for everyone involved in the project to witness the power of this approach in action. The project enjoyed quick success and was subsequently funded by Bosch.
Peter brought with him a rich background of knowledge and experience in the design and direction of R&D projects for many different kinds of organizations, including NASA, NIST, the US Navy, industrial firms, private research institutes, and academia. He started his professional career in 1965 with professional and industrial software development at IIT Research Institute (1965 - 1969), where he was responsible for the design of the Experiment Profile Analysis (EPA) technique, a computer simulation which was used by the NASA to determine the most efficient spatial and temporal sequences for remote sensing activities on the Apollo Earth and Lunar orbital missions.
When the Apollo program began to wind down, Peter joined the faculty of the Department of Electrical Engineering and Computer Science at The George Washington University (GWU) in 1970, designing and teaching computer science courses, including artificial intelligence, adaptive learning systems, cognitive science, robotics, and simulation. He was primarily responsible for the development of the graduate degree program in Machine Intelligence and Cognition. It was here that he pioneered the research in Collective Learning Systems Theory. Reflecting the neurophysiological and psychological processes of the mammalian brain, he designed a hierarchical network of learning automata to simulate the architecture and adaptive function of the human cerebral cortex: learning from nature how nature learns.
It was this research that eventually led to the joint American-German project ALANN (Adaptive Learning and Neural Networks) at the FAW in Ulm, Germany, which designed and built the first ALISA system (Adaptive Learning Image and Signal Analysis). ALISA is used today in several learning and classification applications in industry and research, and is currently being expanded by Peter's research group at GWU into a complete hierarchical system for artificial cognition.
Over the years, Peter and I have sat together on many occasions and lamented the fact that so many of our young scientists and engineers leave the universities with a solid foundation of knowledge in their disciplines, but with little idea how to design and conduct real-life R&D projects successfully. Finally, Peter decided to try to do something about this problem. In 1996, he set up a seminar at GWU to teach the Scientific Method for R&D in science and engineering. We also brought this seminar to Bosch Corporate Research in 1997. His lecture notes quickly expanded into a complete set of basic principles and guidelines for designing and conducting effective research or development projects, both in industry and academia. In 1999, his department added a formal course in research methods to its graduate course curriculum, which is now required for all doctoral students.
At that point it seemed only natural that Peter should transform his lecture notes into a book to spread the word to the research and engineering community at large. By compiling his knowledge and experience into a reference book, his intention was to provide a complete and consistent methodology for scientists and engineers, project managers, and administrators to plan and run research and development projects. In 1999 Academic Press indicated its interest in just such a manuscript, and voila!, here it is. The book covers project organization, knowledge representation, and a modernized version of the classical Scientific Method, which has been adapted for the special requirements of research and development in the 21st century. The question facing Peter, however, was how to present this complex material in a reasonably linear fashion, while avoiding the sterility of an IF-THEN-ELSE cookbook approach and a dry and overly academic presentation. It is a bit of a mystery to me exactly how he managed to do this, but in fact he did. He has succeeded in coming up with a smart, readable book that guides real-world practitioners toward the improvement of their daily R&D activities. It provides a clear perspective for people charged with the acquisition, organization, and application of knowledge, devices, and effects, which is the essence of research and development in engineering and science. And it is also a well-structured textbook for an academic course.
A wealth of examples leads the readers to the underlying reasons for failures and successes in the performance of their projects. Carefully constructed summaries, lists of tips, tables, and figures give the reader brief answers about how to avoid common pitfalls and how to do things right. Amusing illustrations catch the reader's eye, add spice to the experience, and serve as convenient bookmarks and reminders for special problems and messages. Peter not only explains the subjects, but also puts the information in its historical and cultural context, making it part of something larger, an essential perspective in the modern multidisciplinary world.
Students will find well-defined and useful exercises to harden their theoretical knowledge, and professionals can use it as a handbook and a lexicon for answering questions that arise during their daily R&D activities. Project managers will learn how to design a project task tree and a milestone chart in evolving stages, using pilot tasks to explore new ideas before scarce resources are committed to formal investigations. Administrators will learn that every project needs a management reserve, an appropriate tool set, and the allocation of adequate time and money to acquire new tools and maintain the project environment. Scientists and engineers will learn about the categories and types of knowledge, arranged in a useful taxonomy.
From an industrial point of view, the most important guidelines in this book provide the means to optimally conduct research and development in terms of effectiveness and efficiency. The fundamental paradigm to accomplish this, the Scientific Method, proceeds from an analysis of the problem and precise statements of the governing propositions, to a clear formulation of the project objectives, to the creation of solutions with well-defined goals and hypotheses, to the design and conduct of rigorously controlled experiments, to the generation of reproducible conclusions or producible competitive products of the required quality. It is clearly acknowledged that research results must be validated by formal peer review within the scientific community, while industrial products must ultimately prove themselves in the marketplace.
Peter's book presents a complete and consistent methodology for research and development in science and engineering. Hopefully, it will encourage institutions of higher learning to include courses in methodology as part of their required curricula, with Peter's book at the heart of their efforts. Read, enjoy, and learn!
Vice President, Corporate Research and Development, Robert Bosch GmbH