Citizen Sensing

The picture is a still from the film ‘Healthy People, Healthy Environment: Integrated Development in Tanzania,’ produced by ECSP.

Along the northern coast of Tanzania, a series of innovative development projects are combining conservation efforts with health and livelihood interventions. We meet Rukia, Mahija, and Fidea from the Pangani and Bagamoyo districts who demonstrate how these “population, health, and environment” projects are improving their lives, their environment, and their community. Community members are some of the most important tools of the PHE approach – local villagers become peer educators who can discuss the linkages of conservation and health with their neighbors and help spread the use of interventions such as clean cookstoves, sustainable seaweed farming, and modern reproductive health services.

Three billion people depend on polluting, open fires or inefficient stoves to cook their food according to the Clean Cooking Alliance factsheet. Social workers are involved on multiple fronts such training school children and community members on alternative, improved cookstoves to avoid challenges of traditional ones such as exposure to toxic fumes. Citing the Clean Cooking Alliance, “traditional wood-fired cookstoves and open fires emit small particles, carbon monoxide, and other noxious fumes that are up to 100 times higher than the recommended limits set by WHO, and in some settings, considerably higher.”

Achieving universal access to clean and modern cooking fuels, technologies and services is a matter of public policy. In fact it is tied to SDG 7 which sees clean cooking as essential to addressing energy poverty and ensuring sustainable energy security for billions of people. Thank the hundreds of clean cooking angels who are at the frontline of this important battle against the “killer in the kitchen.”

There is another name for individuals at the interface between communities, citizens and government policy.

Street Level Bureaucrats

Street level bureaucrats are the teachers, police officers, social workers and other professionals who interact directly with citizens on behalf of the state. The term was introduced in the late seventies and early eighties by Professor Michael Lipsky, now a Senior Program Director at Demos, a public policy research and advocacy organization. In his book, Lipsky addresses the dilemmas of individuals working in Public Service and shed light on the significance of these individuals in the public policy-making process.

Source: United States Studies Centre

Sticking to the cookstove example, there are organizations such as Nexleaf and Berkeley Air who devote their talent in embedded systems development and data analytics to help street level bureaucrats in their work. Nexleaf has developed a solution called StoveTrace over the past 6 years as part of Project Surya. It has been installed in over 700 households across more than 30 villages in India. Similarly, Berkeley Air has developed SUMS and SUMSarizer, a open source software to process and visualize the data collected by the data loggers.

Street-level bureaucrats occupy a particular position which they owe to discretion, or power to decide the type, quantity, and quality of sanctions and rewards during policy implementation.

Street Level Bureaucracy in the Smart City

Major technology organizations including the likes of Alphabet, IBM and Microsoft have long been engaged in programs to apply their knowhow in Web-scale services to E-Government and Smart Cities. IBM had its Intelligent Operation Center for Smarter Cities with its reference deployment in Rio de Janeiro. Alphabet, or more specifically Sidewalk Labs, has more recently been engaged with Toronto’s Eastern Waterfront community in an urban planning endeavor which still raises concerns among stakeholders as they all try to align their interests. These example are somewhat particular, first-of-a-kind and likely belong to the early days of urban digital transformations as highlighted in this BBC episode of The Inquiry, “Are smart cities dumb?”

Tech-driven Smart City initiatives have also been early adopters of advances in applied Machine Learning. Think of the use of mobile network data for urban planning purposes or predictive maintenance models for urban water systems. Alphabet is a major contributor of advances in AI and machine learning and its organizations such as Google, YouTube or Sidewalk Labs naturally benefit from them. A recent by paper Alkhatib and Bernstein (pdf) explores the analogy between discretion as we know it from street level bureaucrats and that of algorithms automatic making decisions for a group of users. With the current state-of-the-art in machine learning, street level algorithms lack reflexive ability to re-define the decision boundary before making its decision.

Technology has long been identified as a potential challenge to street level bureaucrats. With e-government, citizens can obtain some services online, without any or limited assistance. This trend has progressively changed the relationship of street level bureaucrats and citizens: shift from the application of rules to citizens, case-by-case, to co-design and curation of web and mobile applications and processes devised in conjunction with software engineers, UX dessigners, legal experts, and perhaps IoT experts when software is embedded in the physical world.

Other Ways, Other Scales

There’s clearly no place comparing Sidewalk Labs and IBM with Nexleaf or Berkeley Air mentioned earlier. While technology has the ability to reshuffle discretionary power, some organizations, by having a smaller more narrow focus, are able to make more sustainable contributions. Ben Berkowitz’s SeeClickFix has celebrated its 10 year anniversary in 2018. SeeClickFix takes on part of 311 services through its digital platform and mobile application. In his post, Berkowitz acknowledges a positive evolution in the perception of the platform by street level bureaucrats as they were able to leverage it for stronger communication within their respective communities.

Another encouraging example is the Bristol approach to Citizen Sensing. Bristol is a city and county in South West England with a population of over 459 thousand that has put its citizens at the heart of a process of using old and new information, technology, and design thinking to address its challenges and achive change. Bristol also places an emphasis on the notion of “commons” or resources that are not owned by anyone in particular. A clear departure from the top-down Smart City initiatives powered by the tech giants that has earned Bristol an URBACT award in 2017.

Ground Truth Warranty

On this blog, I comment frequently on the use of mobile applications and sensors to collect data about a particular situation. I refer to soft sensing in the case of SeeClickFix where we use our five senses and report a data point. Hard sensing applies to the noise or air quality electronic sensor which automatically posts data at regular time intervals. Both amount to collecting ground truth evidence e.g. for an environmental problem. These are resolutely participatory techniques. Top-down Smart City initiatives harvest data in different ways, raising the questions of transparency and consent. Both approaches are straightforward applications of technology that equally suffer from the algorithmic limitations prevalent to the street-level algorithms mentioned in Alkhatib and Bernstein.

Harvesting ground truth data requires a safety mechanism of sorts. It seems to me that we should aim for a balanced use of technology that does not reduce discretion of the street level bureaucracy to routine or system parametrization. Algorithms can likely be adapted to isolate edge cases and defer to humans who can then apply their experience and judgement to the situation. This reflexivity provides a warranty in the application of policy augmented by technology.

Citizens lack discretion by definition while street level bureaucrats lack the time and manpower to monitor the evolution of situations in-between case visits, as we move towards continuous feedback experiences. With the help of street level bureaucrats, citizens participating in data collection to back their case also avoid inappropriate use of technology and inappropriate harassment of community members.

I had not come across this term before and the potential challenges that can occur across diverse professions working on the front-lines of policy implementation and hope this idea of citizen sensing does not add to the burden but on the contrary acts as enabler. These are after all key members of our community who likely know a thing or two about sensing to make our lives better…