Zahavi bio- 1 -December 6, 2005

From GABRIEL ROTH

4815 Falstone Avenue, Chevy Chase, Maryland, 20815

Phone: (301) 656-6094 email: Fax: (202) 318-2431

December 6th, 2005

YACOV ZAHAVI

Biographical note

Yacov Zahavi was born in Petach Tikvah, in what is now Israel, on July 13, 1926. He graduated as an engineer from the Technion in Haifa in 1951, and served as a graduate intern at the UK Road Research Laboratory in 1951/52. He pursued further traffic studies at the Technical University of Delft, obtaining his doctorate (Sc.D.) in 1973. His doctoral thesis topic was ÒSynthesizing a Transportation Study by the IN ProcedureÓ.

From 1953 to 1954 Zahavi worked for the Tel Aviv Department of Transportation, In December 1954, he joined with Yacov Kolin, to establish the consulting firm of Kolin and Zahavi. In that period he also lectured on traffic studies at the Technion and at the Hebrew University in Jerusalem.

But ZahaviÕs transport interests were worldwide, and he did not find enough scope for them in Israel. In 1969 he left Kolin & Zahavi and joined the staff of Wilbur Smith, for whom he worked (in Columbia and in London) until 1973. In that period he obtained consulting assignments in Washington DC with the US Department of Transportation (DOT) and the World Bank. In 1975, at the request of officials in the USDOT, he moved his residence to the Washington DC area. Much of his work in Washington was done for the DOT (Systems Analysis Division, Research and Special projects Administration); for the World Bank (Staff Working Paper 230, ÒTravel Characteristics in Cities of Developing and Developed CountriesÓ); and the Federal Government of Germany. In addition to his consulting work, he taught, as an assistant professor, at Johns Hopkins University.

Zahavi died on December 20, 1983, following heart failure, while attending a conference in Amsterdam.

In the transportation profession Zahavi is most commonly remembered for his identification of the Òtraveltime budgetÓ, which he first defined as the time spent by an average car on an urban road network. Later he found it more productive to work with the daily Òtraveltime budgetÓ of a ÒtravelerÓ using all available travel modes. His studies convinced him that the traveltime budget was stable and predictable, but not necessarily constant. Nevertheless, Zahavi is often associated with the simplistic idea that average travel time in urban areas has to be ÒconstantÓ at all times and in all places at about one hour a day.

The identification of stable daily travel time ÒbudgetsÓ brought home to Zahavi the importance of travel speeds. He realized that travelers did not save time as a result of increases in travel speed, but that they applied the time saved from some trips for additional travel. This led him to explain the resistance of travelers to switch from private cars to slower public transport modes. This resistance, described by too many as a Òlove affair with the motor carÓ, he explained as unwillingness to lose trips, because of the activities associated with them. This led him to conclude that public transport could attract passengers from private cars only by providing higher speeds, as it does for many trips in cities such as London, New York and Paris.

The idea that travelers use the fastest travel modes they can afford led Zahavi to identify and develop the concept of a stable ÒTravel Money BudgetÓ which he found to be about one eighth of household income. He went on to develop a predictive model for urban travel, the Unified Mechanism of TravelÓ (UMOT) based on the proposition that travelers tend to maximize distance traveled within their constraints of time and money. The model was called ÒUnifiedÓ because its components (e.g. trip rates and trip distances) were influenced by one another, and determined by a process of iteration. The UMOT model was assessed by the UK Transport and Road Research Laboratory (where Zahavi had served as an intern thirty years earlier) in the 1983 TRRL Supplementary Report 799 ÒUrban Transport modelling with fixed budgets (An evaluation of the UMOT process)Ó. The original model work involved only daily travel, but a later version also determined hourly travel, thus enabling peak-period conditions to be predicted. The model was comparatively simple and ran on an Apple II+, IIC, or IIe computer.

Zahavi regarded urban structure as closely connected to transport, for example trip length being associated with city size. He was anxious to develop models connecting travel and urban structure and regarded Travel Probability Fields as an appropriate tool. A report on this approach was prepared with Martin Beckman and Thomas Golob and published in 1981 by the USDOT as Report No. DOT/RSPA/DPB-10/7 ÒThe ÔUMOTÕ/Urban InteractionsÓ. At the time of his death, Zahavi was working to expand UMOT to enable it to incorporate further relationships with urban structure.