Operations Research at Bendix Corporation in the 1950s and 1960s
Bendix Corporation was formally established in 1924, initially as a vehicular brakes manufacturer. Five years later, Victor Bendix, the company’s eponymous founder, formed a separate aviation division which became the catalyst for the company’s operations research efforts. This included work in cost analysis, mathematical modeling, control theory, and computer science. In addition to their internal advancements, Bendix researchers participated in the greater and quickly growing operations research and management sciences community of the mid-twentieth century. As a significant manufacturer of household (e.g. TV components and washing machines) and military (e.g. army radios) electronics, the company also built and sold computing tools for its own staff and other OR users across the country.
There were numerous Bendix offices engaged in operations research projects. The primary of these was the Bendix Aviation Systems Division, headquartered in Ann Arbor, MI. It opened on August 18, 1958, initially employing one hundred forty engineers and scientists (Ann Arbor News 1958). In the mid-1960s, the division ran a hiring campaign in Operations Research, looking for senior systems analysts with bachelor degrees who were “knowledgeable in cost and effectiveness analysis, total system concept evaluation, mathematical modeling, and/or Reconnaissance and Inspection Systems (Bendix Corporation 1965).” Other groups within the company had operations researchers during the period, including the Manufacturing Research Department of Bendix Radio (an acquired subsidiary) and Aviation Electronics Division in Baltimore, Bendix Aviation’s Red Bank Division in Eatontown, New Jersey (known primarily for electronics power machinery), the Bendix Product Division in South Bend, Indiana (home to the company’s original headquarters), and Bendix Aviation Research Laboratory Division in Detroit.
Like many institutions in the era of early operations research, Bendix hired people from a diverse array of backgrounds. F. Gordon Barry of Bendix Radio, for instance, was a champion table tennis player who traveled with the Harlem Globetrotters prior to working with the U.S. Signal Corps Repair Shop during World War II. He went on to study electrical and administrative engineering at Johns Hopkins University. In his twelve years with Bendix Radio, Barry developed and taught a management development course centered around industry case studies. He later wrote an article about the course at the request of the Harvard Business Review (Barry 2001). Influential control theorist Yu-Chi Ho started his career at the Detroit research labs, working on the invention and commercial development of numerically controlled machine tools under E. Calvin Johnson in 1955-1958. Ho and Johnson’s work was featured in LIFE magazine and resulted in four patents (Ho 2018).
Like other key players at the time, Bendix was open to collaboration with academic institutions. Hartsel McClain of Bendix Baltimore co-wrote a study of optimum assembly runs with Edwin S. Mills of Johns Hopkins University. The model includes “as decision variables both the length of the run and the rate of assembly [formulated] so that these decisions can be made simultaneous in a routine, clerical way.” They also incorporated slight modifications that made it possible to consider “more complicate material-ordering, production runs policies (Mills and McClain 1961).” In the resulting 1961 article published in Operations Research, the authors thanked W. A. MacCrehan of the Bendix Aviation Electronics Division for supporting and undertaking the initial implementation of the model.
The growth of OR at Bendix corresponded with the growth of the larger OR community. In 1953, a single Bendix employee, Walter S. Hoyt, was among the initial members of the Operations Research Society of America (ORSA) (Operations Research Society of America 1953). By 1957 that number had grown to nine: Archie Blake and R. Stanley LaValle from the Systems Analysis Division, Robert G. Fitsgibbons and Richard E. Wiffen of the Bendix Products Division, Ernest W. Karlin, Robert R. Meijer, and William J. O’Leary of the Red Bank Division, Richard W. Blue of Aviation Research Labs, and F. Gordon Barry from Bendix Radio (Operations Research Society of America 1957). ORSA was not the only professional organization to draw Bendix participation. W.A. MacCrehan presented “Commercial Aspects of Guaranteed Reliability” at the Institute of Radio Engineer’s Fourth National Symposium on Reliability and Quality Control in Washington (Institute of Radio Engineers 1958). And in the realm of professional management science, Bendix’s Vic Miller co-led a discussion on cost reduction with the Corning-Elmira Chapter (upstate New York) of The Institute of Management Science (Juggenheimer 1971).
Among the many Bendix electronics products, the G-series of computers not only proved its worth in sales, but as a useful tool for the company’s internal operations researchers. For instance, Mills and McClain performed their aforementioned computations on a Bendix G-15 (Mills & McClain 1961), a vacuum-tube computer introduced in 1956. The G-15 was developed by Harry Huskey, a former colleague of Alan Turing, and inspired by the Automatic Computing Engine (ACE). Huskey worked with Turing on the Pilot ACE computer during a year-long stint at the National Physical Laboratory in the United Kingdom. Around the same time, he worked on the EDVAC and SEAC computer projects and managed the design and production of the Standards Western Automatic Computer (SWAC) at the National Bureau of Standards in Los Angeles. The G-15 computer may hold the claim of initiating the age of the personal computer. Huskey designed the G-15 so that it could be used by a single person – highlighting this capability by keeping one at his home. Another member of the G-15 engineering team, David C. Evans, became well known later for his work in computer graphics with Ivan Sutherland.
In 1961, Bendix unveiled a successor to the G-15, the Bendix G-20. It was twice as large and comprised a high-speed central processor, a control debugger, and matched input-output units. A review by the Office of Naval Research’s Mathematical Science Division deemed the system “well suited” for scientific computing, data processing, and business computing, and described its “excellent facilities for automatic compilation of programs from algebraic and business-oriented languages (Goldstein, ed. 1960). The G-20 proved useful for outside operations researchers as well, such as Abraham Charnes and W. W. Cooper of Carnegie Tech, who used the university’s computer in their critical path analyses (Charnes, et al. 1964). The G-20 did not have the opportunity to achieve the commercial successes of its predecessor, as Bendix sold its computer division to Control Data Corporation in 1963, prematurely terminating the product line.
The sale of the computer division to Control Data marked a trend that would define Bendix in the following decades. The 1970s and 1980s saw a public dilution of the Bendix corporate identity as the company engaged in numerous back-and-forth mergers and acquisitions. This process culminated in 1983 when Allied Corporation came in as a “white knight” and purchased the company on friendly terms. The washing machines and other well-known product lines kept the Bendix name until 1999, when Allied purchased Honeywell and rolled all appliances and avionics under the latter’s brand.
Links and References
Ann Arbor News (1958) Bendix Aviation Corporation’s Systems Division plant at 3400 Plymouth Rd, August 1958 [photograph]. Ann Arbor District Library. Retrieved from: https://aadl.org/node/366595.
Barry, F. Gordon (2001, November 6) Obituary. [BRF Personnel File] Bendix Radio Foundation (3.01.3.02., F.002.02) Baltimore Museum of Industry, Baltimore, MD.
Bendix Corporation (1965) [advertisement] Operations Research, 13(2): xx.
Bendix Radio Foundation (2017) Bendix Radio History. Web Page. Bendix Radio Foundation. Retrieved from: http://www.bendixradiofoundation.com/hist_main.htm.
Charnes, A., Cooper, W.W., & G. L. Thompson (1964) Critical Path Analyses via Chance Constrained and Stochastic Programming. Operations Research, 12(3): 460-470.
Goldstein, G.D., ed. (1960) Computers and Data Processors, North America: Bendix Computer Div. of Bendix Aviation, G-20, Los Angeles, CA. Office of Naval Research, Mathematical Science Division Digital Computer Newsletter, 12(2): 1-6.
Hansen, Charles (2019) A Brief History of Bendix Red Bank Tubes. KCK Media Corp: Middletown, DE.
Ho, Yu-Chi (2018, June 30) Numerically Controlled Machine Tools [Blog post]. Retrieved from: http://blog.sciencenet.cn/home.php?mod=space&uid=1565&do=blog&quickforward=1&id=1121514.
Institute of Radio Engineers (1958) Proceedings, Fourth National Symposium on Reliability & Quality Control in Electronics, Washington D.C., January 6-8, 1958. Institute of Radio Engineers, New York, NY.
Juggenheimer, R. W. (1971) College and Chapter News. Interfaces, 1(4): 70-83.
Mills, E. S., & H. G. McClain (1961) A Study of Optimum Assembly Runs. Operations Research, 9(1): 30-38.
Operations Research Society of America (1953) Members of the Society, July 1, 1953. Journal of the Operations Research Society of America, 1(4): 8-16.
Operations Research Society of America (1957) Members of the Society. Operations Research¸ 5(6): 8-66.