Click for homepage
R.M.M. Oberman
Professor, scientist and inventor

Prof. Dr. Ir. Roelof Maarten Marie Oberman (Leiden, 12 Juni 1910 – Wageningen 19 March 1989) (age 78) was a Dutch academic who worked as a developer for the Dr. Neher Laboratory in Leid­schendam (Netherlands) – at the time the prestigious research department of the Dutch state-owned telecom monopolist PTT – and later also as a professor at the Technical University of Delft.

After graduating twice from Delft University, 1 Oberman began working for the Dutch state-owned telecom provider PTT in 1935, where he became responsible for the development of the telegraphy network (telex). In 1946 he made the move to the PTT's Dr. Neher Laboratory, which was regarded at the time as a state laboratory.

In his capacity as senior developer at the Neher Lab in Leidschendam (Netherlands), Oberman developed the first Dutch post-war OTT cipher machine, a relay-based mixer known as Colex, at the request of the Dutch Government.

Colex evolved into Ecolex I and Ecolex II, both of which were eventually taken over and produced by Philips Usfa. It marked the start of Philips's crypto developments, which lasted until 2003.

From 1947 to 1958, Oberman worked both for the PTT and as Professor at Delft University.
  
Click to see more

In 1949, at the official commissioning of the Colex cipher machine between five embassies and the Dutch Foreign Office in The Hague, Oberman was decorated as Officier in de Orde van de Nederlandse Leeuw (Officer of the Order of the Dutch Lion) by Prime Minister Willem Drees.

Apart from cipher machines, Oberman also developed an allotment machine for the Dutch State Lottery (Staatsloterij), which was a spin-off from the key tape generator that he had developed for the Colex and Ecolex cipher machines. In 1958, he left the Dr. Neher Lab of the PTT to become full time Professor at Delft University, where he lectured switching technology – at the time the building blocks of telephone switches and cipher machines – until his retirement in 1980.

During his lifetime, Oberman filed more than 90 patents, most of which were registered in multiple countries, including Germany, the UK and the USA. Many of his inventions involved (telephone) switching systems, for which the patents were filed on behalf of the Dutch State or the PTT. Most of the crypto-replated patents however, were filed on behalf of Philips Usfa.

 Full biography

  1. At the time known as Technische Hogeschool Delft.

ANNOUNCEMENT

December 2020 — Currently in preparation is a paper about Professor Oberman and his crypto-related work during the years 1946-1965, which will be presented at the forthcoming HistoCrypt Conference, that will be held in The Netherlands in September 2021 (subject to Corona crisis developments).

People who have background information or (better) documentation about Professor Oberman's crypto-related work, or who have worked with him in the past on the subject, are kindly invited to contact us. Also, if you have a specific question about Professor Oberman's crypto-related work, that could assist the research, we would like to hear from you.

Related items on this website
PTT Colex (predecessor of Ecolex)
EROLET
Ecolex-I
Ecolex-II
Known products
Biography
Oberman graduated in 1932 as Mechanical Engineer (Ir.) at the Technical University of Delft (Netherlands). As there wasn't much work at the time, he was given the opportunity to keep studying, as a result of which he graduated again in 1935 as Electronics Engineer (Ir.).

After his graduation, he found a job as Telegraph Engineer at the PTT district of Den Bosch (Netherlands). A year later, in 1936, he became chief of the technical service department of the PTT's Telegraph Office in Rotterdam. In August 1940 he was transferred to the PTT headquarters in The Hague, where he became responsible for the Telegraphy Department.

Post WWII
After the war, in 1946, Oberman became head of the equipment department at the Dr. Neher Laboratory of the PTT. In 1947, he graduated cum laude at Delft University, and became Doctor of Technical Sciences, with his dissertation The Bridge Marker Key Automatic Switching System.

In May 1947 he was given a teaching assignment at Delft University, and became extraordinary Professor with a chair in Telegraphy and Switched Telephony. In 1958 he became a full time Professor in Switching Technology — the technology behind information processing systems, telegraphy, telephone switches and also behind cipher machines. Until that time (1958) he had worked at the PTT was well as at Delft University. In 1980 he retired at the age of 70.

Cipher machines
During and also due to WWII it had become clear that message communication systems were prone to malicious interception, also known as co-reading or eavesdropping, even if they were cryptographically protected. This happened with the Enigma cipher machine, but also to the Dutch WWII Intelligence Bureau (BI), which suffered great losses because its messages had been read by the enemy. Immediately after the war, in 1946, the Dutch Government therefore asked the PTT whether it was possible to develop a secure telegraph system, that would be immune to cryptanalytic attacks. They did not want to buy such systems from foreign manufacturers, as they were afraid that they might have hidden weaknesses or, worse, a deliberately built-in backdoor.

Colex
The question finally landed on Oberman's desk in 1946/47, as he was responsible for telegraphy matters within the PTT. This led to the development of the Colex machine — Codetelex — which, after two trial version, met the requirement of all parties involved. Colex was a One-Time Tape cipher machine – a practical implementation of the unbreakable One-Time Pad (OTP) for use in telegraphy communications – and consisted of two parts: a mixer and a key generator. When the latter meets the requirements of true randomness, the system can be proven mathematically to be (forever) unbreakable. The first implementation was completely relay-based, due to the post-war lack of electronic components, and was capable of encrypting just 3 characters per second.

The disadvantage of One-Time Tape is the problem of key distribution. Based on the Vernam Cipher, the key has to be at least as long as the message, and can only be used once, to prevent successful cryptanalytic attacks. Furthermore, it cannot be used by multiple parties and must be destroyed after use. A stringent procedure is required for generating and distributing the key tapes, but it was (and still is) the only way to technically ensure that State Secrets are kept secret.

Part of original circuit diagram of the Colex cipher machine. Copyright Crypto Museum.

The mixing part of the Colex (i.e. the mixer) was constructed with approx. 100 electromechanical relays, that mix each letter of the plaintext with a character from the key tape, using an operation that was carried out with a mix and unmix feature. Several years later this became known as the exlusive-OR operation (XOR), also known as modulo-2 addition. If both parties have the same key tape and use it from the same starting position onwards (i.e. synchronised), the recipient has to perform the same XOR operation in order to reveal the plaintext again.

One of the critical parts of the system is the key generator, which Oberman had named Roulette. Facing the lack of electronic parts immediately after the war, he composed a creative solution to obtain and control randomness of the key production in the Colex system. It comprises a 400 kHz oscillator of which the output is divided by 2 in five stages. Each stage produces one of the five bits of the ITA2 telegraph alphabet (Baudot), whilst a relay is used to sample the output at ~225 ms intervals. The uncertainty of the movement of the electromechanical relay is used as the source of randomness (noise). To check whether the noise was evenly distributed, a set of five electromechanical counters were connected to the machine, allowing the user to check whether each of the five output bits appeared on the key tape an (approximate) equal number of times.

Oberman's signature on the original Colex circuit diagram

The Colex was built by a small team under Oberman, and was put into service in on 6 April 1949 by Dutch Prime Minister Willem Drees. It was used for communication between the Foreign Office and the embassies in London (UK), Paris (France), Washington (USA) and Jakarta (Indonesia).

 More about Colex

Ecolex
Colex was succeeded by Ecolex — Electronic Codetelex — in which the electromechanical relays were replaced by thermionic valves (vacuum tubes). This system was later renamed Ecolex I, and was succeeded around 1956 by Ecolex II, 1 in which the valves were replaced by transistors. In the same vein, the Roulette key generator was succeeded by Erolet, short for Electronic Roulette.

As Oberman's employer — PTT — was a service provider and not an equipment manufacturer, the production of cipher machines was not regarded as core business, as a result of which in 1956 or 1957 the production of the Ecolex machines was given to Philips Usfa in Eindhoven (Netherlands), who also took over development.

In 1958, Oberman left the PTT and became full time Professor at Delft University. His major subject was switching technology (logic), of which cryptography increasingly became part, especially when in the 1960s cipher machines became based on electronic shift registers.
  
Philips Ecolex-II with Siemens T-send-77b on top

In 1963 Philips released the Ecolex IV, but the operating principle had remained unchanged. Typical customers at the time were the Foreign Office, the Department of Defense, NATO and several Danish and German organisations, but the PTT had gradually lost interest in the matter.

On the first page of his book Counting and Counters [II]

Oberman retired from Delft University in 1980 at the age of 70. Of his aprox. 90 patents, about 25% is related to cryptography. A more elaborate article on the Dutch history of cryptographic equipment and its organisation in the period 1945-1965, including many aspects of the Colex machine, is expected in September 2021.

  1. The first version of the Ecolex II was built with valves (Ecolex IIa), whilst a later version (Ecolex IIb) was built with the first generation transistors.

Publications
  1. Digital Circuits for Binary Arithmetic

  2. Counting and Counters
    ISBN 0-333-30512-4. 1981
Further information
Any links shown in red are currently unavailable. If you like the information on this website, why not make a donation?
Crypto Museum. Created: Saturday 10 August 2019. Last changed: Monday, 11 January 2021 - 17:43 CET.
Click for homepage