Douglas T. Ross papers

Douglas T. Ross was born in Canton, China, on December 21, 1929, the son of Robert Malcolm Ross and Margaret Taylor Ross, who were medical missionaries. He spent most of his childhood in Canandaigua, New York, and left to attend Oberlin College in Oberlin, Ohio, where he majored in mathematics. He married the former Patricia Mott, another Oberlin student, during his senior year. He graduated with an A.B. cum laude in math in 1951.

Douglas Ross enrolled at MIT in 1951 with the intention of doing graduate studies in mathematics. He worked as a teaching assistant in the Mathematics Department, and also worked at the Servomechanisms Laboratory in the summer of 1952. He taught himself computer programming on the Whirlwind Computer and his interests gradually shifted from math to computer science. In 1954 he received an M.S. without course specification. Having finished all required course work for a Ph.D. in pure math, he never completed a dissertation because of increasing commitments and responsibilities involving computer research.

The focus of Ross’s research at MIT from 1952 to 1954 was development of generalized programs for computation of autocorrelation functions and power density spectra using Fourier transformation. Another priority from 1953 to 1957 was developing new techniques for large scale data reduction using on-line manual intervention.

Douglas Ross joined the Computer Applications Group (CAG) when it was part of the Servomechanisms Laboratory at MIT in 1952 and was involved with numerical control (N/C) research from its beginning, being a leader of the original N/C development team. N/C work was sponsored by the U.S. Air Force, which continued its backing when CAG became part of MIT's Electronic Systems Laboratory (ESL). He headed CAG at ESL from 1956 to 1969. From 1956 to 1959 he had technical responsibility for the development of APT, the Automatically Programmed Tools research at MIT, funded by the aerospace industries. APT research aimed at automatic programming for numerically controlled machine tools.

In 1959 Douglas Ross became head of the Computer-Aided Design Project (CAD) at CAG. The work on CAD involved research and development in language theory, language design, generalized computer instruction, computer graphics hardware and software, and design applications. After 1962 he focused on AED Language and Compiler development and was active in the AED Cooperative Program at MIT starting in 1964. He taught the first graduate course in software engineering (6.668) in the spring of 1968 at MIT.

In 1969 Douglas Ross, Jorge Rodriguez, and Clare Feldmann (two of his graduate students), left MIT to found SofTech with two others, John Brackett and John Walsh. Early work at the spin-off company involved providing support for industries using AED and providing functional analysis and system design services using the Structured Analysis and Design Technique (SADT). Douglas Ross served at different times as president, chair of the board, principal investigator, and chair emeritus of SofTech. While at SofTech he was (in his words) "responsible for everything," but not really in charge of any particular operation.

Starting in the 1950s, Douglas Ross became interested in PLEX, which he describes as a scientific philosophy concerned with understanding our understanding of the nature of nature. PLEX is considered by Ross to be the central source upon which his other work has been based.

Douglas Ross was an active participant in the American Association for the Advancement of Science (AAAS), the American Ordnance Association, the Association for Computing Machinery (ACM), the IDEF Users' Group, the Institute of Electric and Electronics Engineers (IEEE) Computer Group, the U.S.A. Standards Institute (USASI), and the International Federation for Information Processing (IFIP). He was a member of IFIP Working Groups 2.1 (Algol), 2.3 (Programming Methods), 2.4 (System Implementation Languages), and 5.1 (Discrete Manufacturing). One of his special interests was in developing standards for computer technology. He served as chair of the Panel for the Institute for Computer Sciences and Technology for the National Bureau of Standards (NBS).

Douglas Ross was the author of many articles and reports and was the co-author of Notes on Analog Digital Conversion Techniques. He served on many MIT Committees (including the Executive Committee of the Enterprise Forum), as well as charitable, church, and civic committees. He belonged to the fraternity Sigma Xi. He and his wife had three daughters.

Douglas Ross died January 31, 2007.

SofTech History

While putting the finishing touches on the third version of AED (Public AED) at MIT in 1968, Douglas Ross and his associates at the Electronic Systems Laboratory (ESL) decided that the time had arrived to form a private company that could provide more support to users of AED than could be staged from an academic setting. Many of the companies that had cooperated in the AED Project at MIT needed such support to derive maximum benefits from their initial application of the technology. Douglas Ross, Jorge Rodriguez, Clarence Feldmann, John Brackett, and John Walsh founded SofTech in 1969. It was essentially a spin-off company carrying on work that had been in progress at ESL. Rodriguez and Feldmann had been Ross's graduate students, and Brackett was also a graduate of MIT. Twelve people were on the original payroll.

SofTech had no proprietary right to AED, but its personnel knew how to use AED better than anyone else and marketed these skills. However, the founders quickly discovered that the company needed to broaden the range of applications of their problem solving methods to survive. After 1973, the Structured Analysis and Design Technique (SADT) developed by Ross formed the basis for much of SofTech's work. The company became known for taking complex problems and breaking them down into discrete parts of manageable size by means of SADT. A major reason for SADT's success was the personal interactions established between SofTech personnel and employees at the client company or agency, ensuring that real problems were identified and useful solutions proposed or designed. Other languages and tools were used as well in SofTech's work. Some were built, others were borrowed or adapted. The company created custom software products and provided systems engineering and integration services for government agencies and industrial clients. The overall approach was to exploit computers as tools for making computers more useful for solving problems. The application of languages, systems, and tools to specific problems is illustrated by the Projects section of the collection. These materials are organized by project name and then by proposal number and contract number.

The following list attempts to categorize the type of work done by SofTech on each of the projects included in the collection. In some cases a project appears in more than one category because the nature of the work performed was multifaceted. SofTech developed a language, system, or tool on the Ada, ACOS, CAMIL, GMS, JOVIAL, ORAS, and UFCS projects. The company analyzed existing systems on the AEDCAP, AFIRMS, ICAM/AFCAM, and ULCS projects. It performed verification and validation operations on the EAR, MAIDS, PLSS, THLL, and ULCS projects. It adapted a language, system, or tool on the AED for CDC and APT for Boston Digital projects, and on the CIRCAL-2, SADT for DEC, OSC DAIS, PL/1 for Time, PLEXUS, and THLL projects. It coordinated an effort on the ICAM and GMS projects, created a spreadsheet operation with ORAS for General Motors, and set up training operations on the SADT for ITT and TRADOC projects.

In 1981 SofTech went public. The 1980s were a period of acquisition and expansion, which gave SofTech additional business in such areas as supercomputer software systems, automatic testing equipment, and microsystems. By 1988 SofTech had 700 employees and $49,000,000 in annual sales. Branch offices around the country supplemented the activities of the main office in Waltham, Massachusetts. Operational divisions of the company in the eighties included the Industrial Systems Division, the Federal Systems Group, the Government Systems Group, and the System Sciences Group. Activities included systems engineering and integration; verification and validation; computer-aided instruction, design, and manufacturing; testing; software support for users; and, in the mid-1980s, original equipment manufacturing.

SofTech probably made its biggest impact while working on the Air Force's Integrated Computer-Aided Manufacturing Program (ICAM), developing a standard model or architecture of aerospace manufacturing.

One of SofTech's major ventures in the 1980s started in June 1979 when its San Diego branch office became SofTech Microsystems (SMS), a subsidiary established for the purpose of directing activities relating to the newly developed UCSD-p system. The University of California had decided to spin off its microsystems development to avoid tax complications and chose SofTech. SMS became involved in a highly competitive market for microcomputer operating systems, and a joint effort launched by IBM and Microsoft left SMS's UCSD-p struggling. SMS went into product development, making UCSD-p for Macintosh for about two years. The subsidiary was sold to Pecan after several years of declining revenue.

SofTech acquired Compass in 1988. Compass was a service business specializing in the conversion of engineering for customers who needed software portability.

Douglas Ross withdrew from SofTech in 1993. SofTech still maintains operations in Waltham, Massachusetts, although a number of its significant functions have been transferred to Grand Rapids, Michigan.