The Self-Driving Subway Train
And why it has and continues to work well after more than a decade in Japan.
The first subway train not driven by humans
The Sendai Subway 1000 series is an electric train type operating on the Sendai Subway Namboku line in Sendai, Japan, since 1987. The 1000 series was the world's first train type to use Fuzzy Logic to control its speed. In comparison to human-controlled acceleration, the 1000 series is 10% more energy-efficient. Due to its success, the train was the recipient of the 28th Laurel Prize (a Japanese award dedicated to railway vehicles).
Since the application of Fuzzy Logic in the Sendai subway train, many technologies in Japan followed the same path. According to Harpreet Singh, designers and engineers have incorporated Fuzzy Logic in numerous modalities such as “facial pattern recognition, air conditioners, washing machines, vacuum cleaners, antiskid braking systems, transmission systems, (…) unmanned helicopters, (…) weather forecasting systems, models for new product pricing or project risk assessment, medical diagnosis and treatment plans, and stock trading.”
Let’s see how it works.
How Fuzzy Logic works
Fuzzy Logic helps computers make determinations despite the vagueness and uncertainty surrounding complex decisions because it works with statistics and not only with “true or false” terms. This means some statements in Fuzzy Logic can be closer to the truth, and not only “true or false”.
According to Fuzzy reasoning, a given portion of water, for example, can be, let us say, 83% drinkable (which is better than 73% drinkable water) although both would still be adequate to drink. This way of reasoning works just fine in speed controllers because it helps to indicate when the engine should accelerate more or accelerate less, and not only “go or stop.”
A good day-to-day application of this kind of reasoning, for example, can be used in the kitchen when you are cooking pasta. After the water used to cook the pasta reaches a certain temperature you can turn off the fire and wait for the remaining heat (that is in the water) to finish cooking the pasta (since you can cook pasta with low-level heated water too).
You, for example, don’t need to overheat the water as it’s usually done since there is a certain amount of heat needed for the task. However, it is, of course, difficult to be precise in this economizing task (for it depends on many factors, such as the weather in the kitchen, the quality of the water, the pan you are using, and many other details), but a very attentive Fuzzy machine would certainly do it very well.
The first time Fuzzy Logic was used in speed controllers
Fuzzy Logic was first implemented in speed controllers in 1973 at the University of London. Professor Mamdani and his student were trying to stabilize the speed of a small steam engine the student had built, but the engine speed would either overshoot the target speed or arrive at the target speed after a series of oscillations, or the speed control would be too sluggish, taking too long for the speed to arrive at the desired setting.
According to Samuel E. Kisko, “the machine was essentially not doing its job well because its understanding of its environment was poor.” Mamdani, fortunately, had read of a control method proposed by Dr. Lotfi Zadeh (the inventor of Fuzzy Logic), head of the electrical engineering department at the University of California at Berkeley, in the United States. So Mamdani experimented with Fuzzy Logic in his engine controller and it worked very well. The engine reached its target in a third of the time with no oscillations in acceleration.
