Tommy Flowers

Designer of the Colossus code-breaking computer

• Born: December 22, 1905
• Birthplace: London, England
• Died: October 28, 1998
• Place of death: London, England

Primary Company/Organization: Government Code and Cypher School (United Kingdom)

Introduction

Tommy Flowers was an English electronics engineer recruited by computer scientist Alan Turing to help with the code-breaking efforts of the British government during World War II, building on Turing's work breaking the German code known as Enigma. In response, Flowers designed the Colossus code-breaking computer, which decrypted Nazi communiques critical to planning the 1944 D-day landings by the Allies on the beaches of Normandy. He later continued the work he had begun before the war on electronic telephone exchanges, publishing a book on the subject in 1976.

Early Life

Thomas Harold “Tommy” Flowers was born the son of a bricklayer's family on December 22, 1905, in the East End of London, England. As a young man, he worked as an apprentice mechanical engineer at the Royal Arsenal, Woolwich, in southeast London, which in the internecine period was constructing steam locomotives as well as armaments. By night he took courses in electrical engineering, and at age twenty-one he took a job with the General Post Office, working in telecommunications research. In the 1930s, he researched telephone exchanges and the use of electronics, planning a fully electronic telephone exchange system that was interrupted by World War II.

Life's Work

Computer scientist and mathematician Alan Turing had introduced several enduring concepts to computer science and became one of the most prominent participants in Britain's code-breaking efforts in World War II, work he had begun in 1938. Although the Enigma machine produced the most famous of the German codes that were broken, its codes were neither the last nor the most difficult. After the primary work on Enigma was completed, Turing turned to recruiting more engineers and mathematicians for code-breaking work, among whom was Flowers.

Turing recruited Flowers to work on the bombe, a tool Turing had designed in 1939 at the Government Code and Cypher School at Bletchley Park, to help with the Enigma decryption efforts. Bletchley Park, now the home of the National Museum of Computing, is a Buckinghamshire estate that became the home to Ultra, the code name for the main Allied code-breaking efforts. The Government Code and Cypher School housed at Bletchley Park, wartime employer of both Turing and Flowers, had been formed in the internecine period to replace the separate signals intelligence agencies of Britain's army and navy. Although the German codes were the most famous that were deciphered, when Flowers was recruited codes and ciphers of twenty-six countries were being studied, in conjunction with the Far East Combined Bureau in Singapore.

Turing's bombe was the primary machine used in Enigma code breaking. Enigma was not a single code but a method for creating codes: The Germans used a rotor machine with a system of rotors, each of which had substitution characters on it. Each time a wheel turned to encrypt a letter, the rest of the rotors would turn, making for a complex polyalphabetical substitution cipher. Some of these codes were nearly impossible to decrypt using methods available at the time. The bombe, which had to be fed a crib (a fragment of possible decrypted text), generated likely rotor machine settings to assist in decryption efforts. The Lorenz machines, meanwhile, were based on a system similar to that used by the Enigma, but instead of the alphabet, Lorenz machines used the larger Baudot character set, which like today's ASCII included letters, numbers, and special characters. The Baudot character set was used in teletypewriters, and the Lorenz machines were used by the Germans to encrypt messages sent on such teletypewriters. It was these codes Flowers worked to decode, following in Turing's footsteps, on a team led by mathematician Max Newman.

What he proposed to do was build an electronic computer, the Colossus, in order to break the code. The government was unpersuaded; the Colossus design called for nearly two thousand vacuum tubes, more than ten times as many as had been used in a single device before. Flowers was accustomed to working with British telephone systems, which were just as complicated, and was convinced the computer would work. He staked some of his own money to begin the project, working at the Post Office Research Station at Dollis Hill, London, with his own team. It took eleven months, from February to December 1943, to build the Colossus, and after a successful test it was sent to Bletchley Park for the code breakers to use, while Flowers and his team proceeded to design and build an improved model, the Colossus Mark 2, which was finished in 1944, used twenty-four hundred vacuum tubes, and was five times faster than the Mark 1. The original Colossus Mark 1 was later converted to a Mark 2 machine. By the end of the war, eleven Colossus computers had been built, greatly speeding cryptanalysis efforts, although parallel computing—in which two Colossus computers worked on different possibilities for solving the same problem—was used only occasionally.

The Colossus was the first programmable electronic digital computer. The first electronic digital computer, the Atanasoff-Berry Computer (ABC), had been built by John Vincent Atanasoff and Clifford E. Berry in the United States before the war, but few people were aware of it. The Colossus was programmable rather than being hardwired for a specific task, like the ABC. It emulated the Germans' rotor machines and generated possible key combinations for use in decrypting Lorenz messages, processing at a speed of five thousand characters per second—faster speeds risked destroying the punched paper tape the computer used. Flowers's telephone exchange experience was invaluable: The reason the government had been skeptical of an electronic device with so many vacuum tubes was that the risk of valve (vacuum tube) failure was higher the more tubes were used. Telephone exchanges were less subject to this problem because their circuits remained turned on all the time; the highest risk of failure was during the power surge that occurred when the computer was initially turned on. The Colossus computers were designed by Flowers to be operated the same way British telephone exchanges were, simply by remaining turned on. They were powered down only when repairs were needed following a malfunction—an expense that at the time only a government could afford.

The Colossus lacked stored programs. Operators adjusted plugs and switches in order to set new programs for an incoming task. It was also not a general-purpose computing machine; like the ABC it had been set up to perform specific kinds of mathematical operations—in this case, the Boolean operations and counting required of cryptanalysis. Like most early computers, it was thus not Turing-complete (computationally universal); the significance of Turing's definition of the computing machine had not yet been realized.

Flowers was never fully compensated for the money he spent on the Colossus, although he was granted £1,000 and made a member of the Order of the British Empire. Because the Colossus work was classified, he was not able to capitalize on it later in order to prove his credentials or the veracity of his ideas; he was denied a loan to build a computer, for instance, because the Bank of England believed he could not make it work, and he was unable to refer them to the near-dozen Colossus computers at Bletchley Park. He instead resumed his telephone exchange work at the Post Office Research Station, completing his all-electronic telephone exchange around 1950 despite the strained postwar British economy. He continued working as a telephone exchange engineer for the next few decades, and in the 1970s, after his role in the World War II code breaking had been declassified, he published Introduction to Exchange Systems, summarizing much of what he had learned.

Personal Life

Flowers married Eileen Margaret Green in 1935, while conducting telephone exchange research for the General Post Office. They had two sons, John and Kenneth. He died in 1998 at the age of ninety-two. The access road at the housing development on the site of his old Post Office Research Station has been named Flowers Centre.

Bibliography

Copeland, B. Jack, ed. Colossus: The Secrets of Bletchley Park's Codebreaking Computers. Rpt. New York: Oxford UP, 2010. Print. Of all the histories of World War II code breaking, this one places the greatest focus on Flowers.

Flowers, Thomas Harold. Introduction to Exchange Systems. New York: Wiley, 1976. Print. Flowers's summary of what he learned and innovated working on the Colossus computers, long delayed by the classified nature of his work.

Hinsley, F. H., and Alan Stripp, eds. Codebreakers: The Inside Story of Bletchley Park. New York: Oxford UP, 2001. Print. A collection of essays by Bletchley Park code breakers.

McKay, Sinclair. The Secret Life of Bletchley Park: The WWII Codebreaking Centre and the Men and Women Who Worked There. London: Aurum, 2011. Print. A history of Britain's World War II code breakers.

Sebag-Montefiore, Hugh. Enigma: The Battle for the Code. New York: Wiley, 2004. Print. A broad history of the Allied effort to break the Enigma code.