NEWS AT SEI

“Texas Instruments had recruited Roger Bate to establish a  software engineering group,” recalled Larry Druffel, who was SEI director and  CEO from 1986 to 1996. “Roger would recruit and train people in good software  engineering practices for TI. And then they often were transferred to other TI  programs. I recall visiting him to discuss TI software engineering practices,  and he chuckled and said, ‘A lot of my engineers get recruited away by other TI  programs. That’s all part of my strategy. I want them to do that, but I fight  them every time, so they won’t catch on.’ That was the way he seeded good  software engineering throughout TI.”

Bate’s later work with the Software Engineering Institute  was instrumental in seeding best practices for software engineering worldwide.  Bate, who was a former chair and member of the SEI Board of Visitors, and a  senior member of the technical staff and, later, a visiting scientist at the  SEI, died March 18, 2009, in McKinney, Texas. He was 86.

Born in 1923 in Denver, Bate began college at Cal Tech as a  chemistry major studying under Linus Pauling in 1941. In 1943, in the middle of  World War II, he enlisted in the U.S. Army Air Corps and transferred to the  U.S. Military Academy at West Point. He graduated in 1947 and spent the next  three years studying as a Rhodes Scholar at the University of Oxford, UK, where  he earned degrees in nuclear physics. From 1951 to 1952, he served with the  10th Engineer Combat Battalion in Korea.

It Is Rocket Science

There is a common expression now, meant to express how  simple something is: “It’s not rocket science.” But Bate’s early work was rocket science, and he proved to be  one of the brightest stars in a galaxy of brilliant lights.

As a young Army captain, he was sent in 1959 to be an  instructor at the U.S. Air Force Academy in Colorado Springs, Colo. His  assignment, he explained later, was to help set up the Department of  Astronautics there. He transferred to the Air Force and became the first  permanent professor of astronautics at the Academy in 1962 and the department’s  first head in 1963. He took leave from the department from 1963 to 1964 to earn  a doctorate in aeronautical and astronautical engineering at Stanford.

Larry Jones, of the SEI’s Research, Technology and System  Solutions Program, was formerly chairman of the Computer Science Department at  the Air Force Academy. “I didn’t know Roger at the Academy because he had  already left before I started there,” said Jones. “But he was legendary. And he  was legendarily brilliant.”

Druffel concurred: “Besides Watts Humphrey of the SEI, I’ve  worked with three other National Medal of Technology winners, and,  intellectually, Roger was in their league.”

“He was smarter than all the rest of us,” agreed Jack  Ferguson, a former faculty member in the Department of Astronautics at the  Academy and the former head of the SEI’s Appraisal Program, now a visiting  scientist at the SEI.  “I taught out of  Roger’s book when I was at the Academy,” he said.

“Roger’s book” is Fundamentals  of Astrodynamics, which Bate co-wrote with Donald Mueller and Jerry  White.  The textbook was first published in  1971 and, astonishingly for a technology text, is still in print and still used  in college courses and by professionals today. Internet reviewers of the book  continue to hold it in high esteem: It is known as BMW (after the authors) by its devotees, and one even noted that he  likes to keep two copies—one in his office and one in his briefcase.

What explains the staying power of a textbook nearly 40  years after publication? “I think the authors nailed it,” explained Ferguson.  “It has the right blend of mathematical rigor yet lack of complexity, so  students can understand it.”

Perhaps it was the need for mathematical rigor that led Bate  into computer science. In the early days of the Academy, explains Ferguson, the  Department of Astronautics had the greatest need for mathematical computing  power. So the department members were the biggest users of the Academy’s  computers and needed to develop the programming standards. As head of the  department, Bate established the computer science program and major at the  Academy, which, at first, were part of the Department of Astronautics.

After holding other Academy positions, including vice-dean  of the faculty, Bate retired from the Air Force in 1973 with the rank of  colonel—in retirement he was later promoted to brigadier general—and left the  Academy for Texas Instruments.

In a nearly 20-year career at TI, Bate formed TI’s Advanced  Software Technology Department and served in the Computer Science Research  Department. Bate’s interest in process improvement started in the late 1970s  when he was leading the Advanced Software Technology Department and was tasked  to solve “the software problem.” This phrase was used in the defense industry  to describe the problems that project managers would cite when trying to  explain why their projects were over budget or late. Working with Edith Martin,  who was the Deputy Under Secretary of Defense for Research and Advanced Technology,  and Druffel, who was then the director of computer systems and software in  Martin’s office, Bate helped to organize a workshop that brought together 300  experts to study the issue. One of the results of that workshop was to validate  the idea for the Software Engineering Institute. Druffel would later become SEI  director.

Bate retired from TI in 1991 as chief computer scientist and  TI fellow.

Joining the SEI

“In those days,” explained Druffel, who is also a former  faculty member of the Air Force Academy’s Astronautics and Computer Science Department,  “retirement at age 65 was mandatory at TI. Roger had joined the SEI Board of  Visitors and was serving as chair when he called me one day, and he said, ‘I  have to retire but I’m not ready to quit working yet.’ So I asked him to come  to the SEI. He stepped down as chair and joined the SEI technical staff as an  adviser to the director’s office.”

Druffel provided budget support for 25 percent of Bate’s  time; other SEI technical programs—if they wanted his expertise—would have to  support the other 75 percent. Within a short time, Bate was 100 percent funded.  “I worried that he was failing retirement and apologized one day to his wife  Madeline who said, ‘It is saving his life to stay professionally active.’”

His work at the SEI included the initiation of the Systems  Engineering Capability Maturity Model (SE-CMM). “The SEI could not have pulled  that together without Roger,” said Druffel, “because he came with a background  in astronautics and aeronautics and TI systems. He was the only one at the SEI  who had the credibility in systems engineering to lead that effort and did it  effectively.”

As Bate was working on the SE-CMM, other Capability Maturity  Models were beginning to grow, said Druffel, who asked Bate to chair a group  charged with evaluating the idea of integrating the CMMs in 1994. This work led  to the Capability Maturity Model Integration (CMMI).

“There was a raging debate at the time,” Druffel recalled,  but he says Bate was the “right person to sort that out.”

“He was an intellectual giant,” Druffel said, “but he was  the kind of person whose contributions have often been behind the scenes  pushing other folks. He just had a way of identifying good people and helping  them mature their ideas and follow through on them. So a lot of his  contributions are of an intellectual nature that don’t get the same kind of  recognition, but they’re no less important.”

Sandra Shrum, co-author of CMMI: Guidelines for Process Integration and Product Improvement, 2nd  Edition, said, “Roger’s official title was CMMI chief architect, but he was  also sometimes the chief diplomat. A lot of highly intelligent and very  passionate people were working on CMMI, and that sometimes led to very  passionate disagreements. Whenever this happened in a meeting or working  session, as soon as the voices died down, everybody would look at Roger and  just expect him to say the right thing—to help us move forward.”

Jones, the computer science head at the Academy, noted the  same quality: “I was on loan for the CMMI project. I got to observe his style  first hand and was struck that a guy who was so legendarily smart was not only  smart, but also accessible and down to earth. His leadership style was very  casual. He would definitely form opinions, but he would also listen to  feedback.”

“I spent two days working with Roger at his house in Plano,”  Jones continued, “and he was just so gracious. He worked in a hands-on  collaborative fashion on the details. We had to work out some thorny  issues—some disagreements on fairly fundamental things. The models we were  trying to merge were sometimes purposefully different. So to have brought  people with differing viewpoints under the tent as part of the solution was a brilliant  strategic management stroke. Then we just let the good work and the wisdom of  the people involved come around on the working level as opposed to the  political level. And Roger was the chief voice of reason in his position.”

The culmination of Bate’s work at the SEI came around the  release of CMMI for Development (CMMI-Dev) V. 1.2. The CMMI Steering Group  decided that a new architecture would help CMMI expand its coverage and extend  into other domains. The revised version employs a new architecture and a new term—constellations—to  describe how components are grouped. That description should not come as a  surprise to anyone who knew Bate.

Bate gave some thought to future needs of CMMI users and how  the architecture could meet those needs. “I thought of CMMI’s collections of  best practices as the stars of process improvement,” he told a writer for the  SEI Annual Report in 2006, “and I pushed the metaphor a little further to call  a collection of model, training, and appraisal components for an area of interest  a constellation.” The metaphor came naturally to the former astronautics  professor.

Summing up the contributions Bate made to CMMI and to the  SEI, William Peterson, director of the Software Engineering Process Management  Program, said, “Roger led the SEI’s process work into the realms of systems engineering  and integrated product and process development. With Roger’s leadership, and  with his collaborators’ significant experience and expertise, we were able to  greatly expand and enhance the improvement opportunities available to the U.S.  defense community and to the worldwide development community through CMMI.  Throughout his life, Roger made a lasting, positive impact wherever he was  involved.”

SEI Director and CEO Paul Nielsen called Bate “a significant  member of the SEI team who was instrumental in the creation of CMMI and its new  framework. He was a friend and mentor to everyone who worked with him.”
   
  Nielsen, an Air Force Academy graduate and a retired Air Force major general,  also noted, “Some of us who attended or taught at the Academy remember Roger as  the head of the Astronautics and Computer Science Department. About a year ago,  as the Astronautics Department celebrated its 50th anniversary, then Secretary  of the Air Force Mike Wynne, who was previously a member of the astronautics  faculty,  promoted Roger to brigadier general  for his contributions—a very rare, but richly deserved honor.”

Bate was an SEI fellow, a fellow of the Association for  Computing Machinery, and a fellow of the Society for Design and Process  Science.

In addition to his wife, he is survived by two brothers, eight  children, 11 grandchildren, and 3 great-grandchildren.