8BackgroundMy first contact to Pr<strong>of</strong>essor Kikuchi was through one <strong>of</strong> his Finnish friends, to whom hehad given the task <strong>of</strong> finding out persons working with fuzzy sets in transportation anddecision making in Finland. At that time, in 1991, I had not done any work in this field,but the questionnaire made me interested in the subject. <strong>In</strong> 1992 I had the opportunity tomeet personally with Pr<strong>of</strong>essor Kikuchi in the WCTR in Lyon, where his paper on theLevel <strong>of</strong> service analysis using fuzzy sets was given a best paper award by the WCTRsociety. I was immediately inspired by the human and friendly interest he showedduring our discussion.This way I got interested in fuzzy sets and neural networks and learned to knowPr<strong>of</strong>essor Kikuchi. With his kind support I then had the possibility to spend a sabbaticalyear at <strong>University</strong> <strong>of</strong> Delaware in 1994-1995 and study the fundamentals <strong>of</strong> fuzzy setsand neural networks in transportation. I also had the possibility to participate in theresearch and teaching together with Pr<strong>of</strong>essor Kikuchi.Research <strong>of</strong> fuzzy signal control<strong>In</strong> spring 1995, still in Delaware, I made a research plan and submitted an application tothe Academy <strong>of</strong> Finland for financing research on fuzzy traffic signal control. TheAcademy gave me a three year partial financing for the project and the financing thencontinued for two more periods until I became the President <strong>of</strong> Helsinki <strong>University</strong> <strong>of</strong>Technology in fall 2003 and had no more time for active research. Pr<strong>of</strong>essor Kikuchiparticipated in the work as an outside expert and visited us for a few times during theproject. He also spent two months with us in Helsinki during his sabbatical in fall 2002.As a part <strong>of</strong> our project, one <strong>of</strong> my doctoral students spent a half year period at the<strong>University</strong> <strong>of</strong> Delaware to make research and to learn more about the fundamentals <strong>of</strong>fuzzy sets.Our research <strong>of</strong> fuzzy signal control started by repeating the classic simulation study <strong>of</strong>Pappis and Mamdani using the HUTSIM-simulation model, a simulation system we hadbeen developing from the beginning <strong>of</strong> 1990ies. The test succeeded and we continuedwith the case <strong>of</strong> pedestrian crossings and then moved on step by step to morecomplicated environments. The idea in the research was to use the HUTSIM simulationmodel as the test environment by developing into the program a fuzzy signal controllerin addition to the basic control methods (fixed time, traffic actuation with the extensionprinciple). This way we could compare the control methods in the same environment. <strong>In</strong>the development <strong>of</strong> the fuzzy signal control logic the HUTSIM traffic flow model wasused as the traffic situation model.The research was not only based on simulation, but resulted also in the construction <strong>of</strong> areal signal controller, which was assembled to an intersection in the Helsinki area for testuse. After some practical problems the controller worked fine and measurements fromthe field could be compared with simulation results. The simulations and field testsshowed that fuzzy control can successfully be used in traffic signals. The development <strong>of</strong>the fuzzy control principles included also public transport priorities and the control <strong>of</strong>pedestrian flows at intersections.The results <strong>of</strong> the development <strong>of</strong> fuzzy signal control and the related developmentwork on the HUTSIM simulation model resulted in two Doctoral Theses, one (1)about the simulation and control system and the other (2) about control principlesand testing. <strong>In</strong> addition to the Doctoral Theses several Master’s theses and a greatnumber <strong>of</strong> international scientific journal and conference papers were publishedrelated to this research. Today the development <strong>of</strong> the fuzzy signal controller is doneby a company named RSM Ltd (Road Safety Management, see www.rsm.ie), whohas been developing the ideas towards a commercial product.Closing remarksThe development <strong>of</strong> the fuzzy signal control at Helsinki <strong>University</strong> <strong>of</strong> Technology isa good example <strong>of</strong> a long term research which started in early 1990ies with thedevelopment <strong>of</strong> the HUTSIM simulation model and widened to the development <strong>of</strong>fuzzy signal control principles and to the construction <strong>of</strong> a real controller tested in thefield. The work which started with the development <strong>of</strong> HUTSIM and continued withfuzzy signal control concentrates today on ITS systems and data fusion in general.The work on fuzzy control was originally inspired by Pr<strong>of</strong>essor Shinya Kikuchi whosupported it with his comments and ideas. He is also the main contributor to the idea<strong>of</strong> Helsinki Summer School on Transportation, an activity we started in 2007 andwhich will be described by Pr<strong>of</strong>essor Tapio Luttinen today.Shinya Kikuchi was a warm and friendly person and a great friend <strong>of</strong> Finland. Hewas open to cooperation and always ready to help others. He was fully committed tohis scientific work, to his teaching and to his students. He was a pioneer in theanalysis <strong>of</strong> uncertainty in transportation systems and in the use <strong>of</strong> fuzzy sets andneural networks in transportation.During the years <strong>of</strong> our cooperation my wife and I became close friends with Shinyaand his wife Laura. To Shinya, Laura was an important background support withoutwhom Shinya’s total and successful commitment to his work had not been possible.So, I want to express my personal condolences to her for the great loss she togetherwith us, the friends and colleagues <strong>of</strong> Shinya, has faced.For more informationHUTSIM and fuzzy signal control are described in the dissertations below.Kosonen I. (1999). HUTSIM - Urban Traffic Simulation and Control Model: Principles andApplications. DSc. thesis. Helsinki <strong>University</strong> <strong>of</strong> Technology, Transportation Engineering,Publication 100. Espoo, Finland. 249 p.Niittymäki, J. (2002). Fuzzy Traffic Signal Control - Principles and Application. DSc. thesis.Helsinki <strong>University</strong> <strong>of</strong> Technology, Transportation Engineering. Publication 103. Espoo. p. 71 +appendixes.More information about the research on fuzzy signal control in general is available from theTransportation Group at Aalto <strong>University</strong> http://civil.aalto.fi/en/research/transportation/.<strong>In</strong>formation about the summer school is available at http://civil.aalto.fi/en/research/transportation/helsinki_summer_school_in_transportation/.Matti Pursula , Pr<strong>of</strong>essorAalto <strong>University</strong>, Finland 9
10It is not certain for me when I met Pr<strong>of</strong>essor Shinya Kikuchi for the first time. It iscertain it was in the month January, but the year is fuzzy, it might be 1996, but 1998 isalso possible. This range <strong>of</strong> possibility is just the way Shinya used to explain theintrinsic limitations <strong>of</strong> humans to handle facts <strong>of</strong> reality. Nothing is more certain thanuncertainty.I still remember how we met for the first time: it was after a presentation at the TRBannual meeting where he presented one <strong>of</strong> the fuzzy models <strong>of</strong> some transportphenomenon. <strong>In</strong> that time I just finished a study on Limits <strong>of</strong> Predictability. I foundthat many models in transport come with quasi accurate results, single values, whilein fact the results are uncertain due to the model itself, the limited representation <strong>of</strong>reality and the <strong>of</strong>ten incomplete and unreliable input for the model calculation. Theresearchers who contributed to my study also showed that several models <strong>of</strong>transport phenomena have a semi chaotic character: small changes <strong>of</strong> inputparameters could have a huge impact on the results.After Shinya gave his presentation at the TRB meeting where I met him, I wasimpressed by the approach he explained. My first reaction was that I had somedoubts about what he told. I challenged him to give more evidence <strong>of</strong> the validity <strong>of</strong>his approach. His reaction was typical for the man I learned more in the yearsafterwards. He just invited me to join him in writing papers, where his approach wasintegrated with the methods I used. His open mind and friendliness made him morethan a colleague and co-author, I consider him as a friend. <strong>In</strong> the years afterwards weworked together on some articles and conference papers. He stimulated me to rethinkan older publication on the removal <strong>of</strong> errors in traffic counts in a network.What I appreciate in the way he worked as a scientist is that he created opportunitiesfor colleagues, students and himself to develop new ideas and to present them todifferent audiences. He organized the Sunday workshops at the TRB meetings wherehe invited several scholars - also me - to present our ideas on the application <strong>of</strong>artificial intelligence in transportation. He was also involved in the organization <strong>of</strong> ayearly summer school in Helsinki, where he invited me to give a contribution in theprogram. It was a pleasure to work with him. I am thankful for the opportunities hecreated for the development andShinya and I met each other regularly, yearly in January at TRB. He invited me tobecome involved in the committee ABJ70 on Artificial <strong>In</strong>telligence. I admired hisenergy to stimulate the members to work together and contribute to the commission.We also met at the INSTR seminar every three years, the international seminar ontransport reliability. Reliability <strong>of</strong> transport was indeed in the core <strong>of</strong> the domain <strong>of</strong>his interest and we wrote a few papers together, for instance one on anxiety <strong>of</strong>travellers generated by uncertainty. That was an important basis for a later studydone by one <strong>of</strong> my PhD students on uncertainty and travel information. He was also<strong>of</strong>ten present at the European Working Group Meetings in Italy and Poland. Itoccurred more <strong>of</strong>ten that we met each other on a conference that we both chose toattend, because we had similar interests. He told me at that time that he wasconcerned about his health.I have read his last paper he submitted to the present TRB meeting. It containsimportant new views on the evaluation <strong>of</strong> transport plans, plenty <strong>of</strong> thoughts forpractitioners and researchers to apply the concepts <strong>of</strong> fuzzy reasoning to evaluation<strong>of</strong> plans.Shinya Kikuchi is for me an inspiring colleague and friend; I enjoyed working withhim, to learn his original ideas on uncertainty. I am thankful for our friendship theopportunities he created for cooperation.Henk van Zuylen, Pr<strong>of</strong>essorDelft <strong>University</strong> <strong>of</strong> Technology, NetherlandsIt is a great honor for me to write about my friend and colleague, Pr<strong>of</strong>essor ShinyaKikuchi, a gentleman and scholar, engineer and planner, philosopher and guide.I had the pleasure <strong>of</strong> knowing Shinya since 1989, when I met him in Yokohama,Japan, at the World Conference on Transportation Research (WCTR). Ever sincethen we met regularly at the ASCE, TRB, WCTR, and the EURO Working GroupConferences. It was a joy to be with him, sharing ideas and thoughts on researchand teaching. <strong>In</strong> the early part <strong>of</strong> 2000, I was on a sabbatical at the <strong>University</strong> <strong>of</strong>Delaware, and this gave me an opportunity to know him even more intimately as acritical thinker and researcher. He was kind and gracious to everybody who camein contact with him. <strong>In</strong> June 2002, we were together in Bari, Italy, when we werebusy with the EURO Conference. With these and scores <strong>of</strong> other meetings, wewrote papers and discussed research work together. I always counted on Shinya'sknowledge and experience and we were in touch constantly on topics such asFuzzy mathematics, technical education, and the philosophy <strong>of</strong> transportation. Thetransportation world has lost a great teacher and researcher.Farewell Shinya, my dear friend and colleague, I will miss you for the rest <strong>of</strong> my lifeC. Jotin Khisty, Emeritus Pr<strong>of</strong>essorIllinois <strong>In</strong>stitute <strong>of</strong> Technology, Chicago11