Today we live in an era in which our phones, watches, household appliances and our homes themselves are getting smart. At the moment, all “smart” means is “internet connected”. This is a bit ironic as when humans connect to today’s internet we do and say things that are the opposite of smart. Hopefully our refrigerators will be better behaved.
Even the cities we live in may soon be smart if Sidewalk Labs, a venture owned by Google parent company Alphabet, has its way. What practical problems might smart cities solve? Sidewalk Labs’ goal is to improve urban infrastructure through technological solutions, tackling issues such as cost of living, transportation and energy use. In a smart city, self-driving buses replace personal vehicles. Robots bring us our mail and take away our garbage. Modular buildings expand, and presumably contract, as families change size. To this end Sidewalk Labs is experimenting in Toronto’s East Bayfront neighborhood, planning to develop a 12-acre waterfront site called Quayside.
What is notable is not so much the robots and self-driving vehicles but sensors everywhere tracking the movement of people, cars, energy, sewage, goods and services throughout the city. The city senses, collects data and adapts in response to the demands of its inhabitants. For example, most traffic lights today are dumb. Who hasn’t experienced staring at a red light for a couple of minutes, waiting even if there is zero cross traffic to justify not crossing? In a smart city, the lights would turn green for you in that situation. The signaling patterns would change in order to keep things moving, based on sensing actual traffic flow. The same sensing and adaptation would apply to how other goods and services are provided. This is smart because rather than allocating resources too much, too soon or both, this can be done closer to just-in-time, minimizing waste. The smart city constantly collects small demand signals, enabling what lean thinkers would call pull-based fulfillment.
The ability to sense and monitor is one of the three fundamental features of a lean operating system. For that matter, it is essential for any system to sustain itself. At the most basic level we perceive our surroundings. We sense hot or cold. Beyond certain thresholds, human bodies respond automatically. We involuntarily shiver to keep warm or pull the hand away to avoid being burned. This is analogous to jidoka within a lean system. At the next level, we sense that we are sensing. We are aware and thoughtful about what we are sensing, how it might affect us and what our options are to influence that result. We see an andon and some combination of standard and non-standard responses. At the highest level of sense-and-monitor is our ability to reflect on the quality of the actions we take, and learn whether our judgments were good or bad. This happens with each turn of the PDCA cycle.
The word “smart” implies a certain quality of intelligence, a shrewdness or ability to arrive at good conclusions, and even to have this third and highest level of sensing awareness. In this regard “sentient” may be a better term than “smart” for cities. Smart cities lacking AI are not smart but rely on human help make judgments on how to adapt based on the sense-and-monitor data. With practice, smart city AI can not only follow pre-set resource adjusting algorithms but alter and even design their own rules. Smart cities will still depend on humans to define “good” and “bad” outcomes, which then set parameters for the algorithms.
We are told that automobiles will soon be self-driving, internet-connected and software-guided, if not smart. It is curious that not only are the horseless carriages that transport us today dumber than their 19th century predecessors, they are not really even autonomously mobile. So-called “automobiles” require constant human intervention, a.k.a. driving. They sense very little other than heat and the levels of fluids and battery. We trigger a controlled explosion, then do our best to prevent the carriage killing anyone. Horses, on the other hand, while not internet-connected, are smart. Horses can sense and monitor their environment. They do not require their rider to constantly drive them or make small adjustments based on road conditions. Horses are perfectly capable of avoiding other teams of horses and even going from point A to B on often-traveled roads. How did it take us a century of “auto” mobiles and a couple of decades of internet to come to grips with this fraud? Hopefully we will retain our horse sense when upgrading to sentient cities