Avoiding the zombie city algorithm

Social scientists and planners have been asking for granulated data for a long time. Clunky, outdated data preserved in outmoded formats makes Indian urbanization one of the most opaque processes in the world. Yet, at a time when new data technologies and new promises of openness are in the air, social scientists and planners appear to be wary. They seem to think that urban life is up in the air like never before. Corporate and political bluster aside, Indian city managers apprehend heightened uncertainties like never before. What explains this dissonance? What do we need to do create a robust urban imagination?

Cities are palimpsests. The postman knows the boundaries of his daily beat. The bus conductor knows the route that he navigates trip after trip. The milk man knows which door to knock morning after morning. The newspaper boy knows which of his rolled up missiles is to be aimed at which balcony week in and week out. The swanky new shopping mall can only be aligned with the 400-year-old city wall. The households in the gated community are reminded of the fact that they are sitting in the foreshore of the 700-year-old lake when rainwater floods their basements—year after year.

Cities are Janus-like—one face is constantly improvising and learning new routines of self-regulation. The other seeks to meticulously create rules that are accepted as legitimate by everyone for the whole to be orchestrated. Planning, regulating and catalyzing change in such places requires data. But it equally requires discretion. It requires intimate knowledge of the local. Change driven exclusively, and undemocratically is a recipe for chaos and misery.

The improvising, self-regulating city operates tactically. It conceals and reveals, dodges and embraces, ducks and thrusts forward, disrupts in minute invisible ways and yet finds its own rhythms. The planned, regulated city operates strategically. It structures, creates opportunities and obstacles. It exercises authority and puts in place infrastructure. Data is created by people who act strategically and tactically. While there can never be a perfect symmetry between the two, they can and should work together, however uneasily, for cities to remain functional. Investments in data technologies without appreciation of this coexistence of different ways of cities create hopelessly entangled legacies of infrastructures and routines.

Take the simple case of bus tracking software. With easy availability of GPS technologies, many schools in our cities are now offering GPS tracking services to parents of children. When the school bus approaches the stop near home, the parent receives a text message. A scroll appears on the local TV screen. The bus is geo-fenced—i.e. when it goes off its ordained route, a deviation alert is recorded and transmitted to everyone. This technology-enabled service is billed as one that saves the parents precious time, makes schools more accountable and children safer. The data is supposed to be useful to ensure that the bus driver is not siphoning off fuel. It is useful to plan school bus routes. So far so good. But what happened to the idea of decent schooling in the neighbourhood? Why are children commuting such long distances through strange, unfamiliar and threatening territories?

This puzzle gets even more intriguing when we consider the fact that till date, not one among the State Transport Undertakings (STUs) which run fleets of thousands of buses in Indian metropolises has fully embraced bus tracking software. Most of them still deploy fleets based on the direct observations reported by controllers and depot managers. Equipped with electronic Ticket Issuing Machines (TIMs), each of the thousands of conductors logs in information about each of the thousands of tickets issued by them in a shift. Most of them still make schedules based on the assumption that running time is a simple function of time and distance, ignoring the rush hour slowdowns. Since a large percentage of crew are contract workers whose monthly income is linked to the number of trips made in a shift, they resort to speeding and skipping scheduled stops. Introduction of data technologies without paying heed to changing work routines and contracting norms has resulted in large amounts of unusable data at one end and disenfranchised bus crews at the other. STUs’ strategy is undermined by the employee tactics.

Take the case of stray dog population in Indian cities. It is the responsibility of the public health wings of municipal authorities to keep the population of stray dogs in Indian cities under check.

In many cities, it varies between five to ten humans per dog. They are required by law and policy to neuter dogs and release them into their old neighbourhoods. Increases in dog population result in higher competition for food and mating and leads to aggressive behaviour and poor health occasionally endangering humans. Yet, more and more municipalities are relying on contract workers, who due to lack of familiarity with neighbourhoods, create war-like conditions on the streets when they catch dogs for sterilization. Hospitals are reporting dog menace near mortuaries and general wards. Dog census, sterilization data, and neighbourhood assessments are all important for creating safe and compassionate neighbourhoods for humans and non-humans—in this case, stray dogs. But equally important are local knowledge, behavioural studies, physical planning and neighbourhood institution building. Smart initiatives that are blind to this lead to programmes like the ones adopted by some smart cities recently—insert microchips in stray dogs, install surveillance systems around mortuaries, hire more data entry workers to monitor the situation.

Government agencies in Indian cities—public utilities, service providers, regulators, local governments—are all virtually under siege from vendors of new technologies.

With large chunks of government work being outsourced, the capacities within these agencies to deal with their day-to-day functions has dramatically come down in the last two decades. The casualty in this is native intelligence. If we approach cities as if all they require is software products to be activated to new life as smart cities, all we will be doing is to pay rent for server stacks for data warehousing.

Smart city initiatives ought to be imagined as interventions into living organisms—not algorithms that can activate doddering zombies. This can be done if we listen out carefully for the urban voices and actors (and that includes dogs and monkeys and other neighbours) that are being systematically ignored.

Anant Maringanti is the director of Hyderabad Urban Lab, a multi-disciplinary urban research institution based in Hyderabad.