- February 11, 2022
“It’s important to talk about resilience as something dynamic. There is no static resilience; it needs to be the resiliency of a process.” -Adrienne Grêt-Regamey, professor and chair of Planning Landscape and Urban Systems at Eidgenössische Technische Hochschule (ETH) Zürich (Swiss Federal Institute of Technology in Zürich)
The Data-Smart team recently read a new report on “Harnessing sensing systems towards urban sustainability transformation.” Given how important the themes of intelligent infrastructure, climate change, and resilient cities are, we reached out to first author Adrienne Grêt-Regamey about the complex relationship between spheres of influence, the role of citizen scientists, sensors, and urban sustainability.
Grêt-Regamey is an environmental scientist and landscape planner who researches how humans understand and change landscapes, and how the landscape changes us. Since 2008 she has been working and teaching at ETH, overseeing extensive lab research that models landscape changes and measures human reactions to these changes. In her acoustic-visualization lab, her team creates models of potential landscapes in virtual reality and measures the physiological changes of participants as they react to these new landscapes. This research informs planning strategies, including behavioral insights into how local governments can utilize incentives and/or regulatory systems to implement greener and more renewable infrastructure.
In “Harnessing Sensor Systems” Grêt-Regamey and her international co-authors develop a conceptual model that combines spheres of influence with active and passive sensing loops in order to frame how urban sensing systems can serve sustainability goals. In order to understand how to utilize sensing systems for increased urban sustainability and resiliency, it’s necessary to break down the model.
Based on the “spheres of influence” idea proposed by University of Oslo professor and climate scientist Karen O’Brien, the authors first break down the difference between active and passive sensing. This was the first big discussion in the researcher group, trying to determine what sensing is and what kind of sensors exist. The two categories of sensors are not solely broken down by technology; rather passive sensors are categorized by low stakeholder engagement and active by high stakeholder involvement. For example, citizen science is active since the person is part of the data and is self-selecting to be an involved stakeholder. A satellite is an example of passive sensing since there’s no active stakeholder participation in the satellite sensing process.
Once the two types of sensors are established, Grêt-Regamey addresses the spheres that these sensing systems exist in. The three spheres are practical, political, and personal with the practical sphere being closest to implementation, followed by the political and then personal. These were arranged based on the idea that the further away from the center, the more difficult it becomes to influence changes. The personal sphere encompasses deeply held beliefs and individual worldviews, which cannot be as easily incentivized or regulated as the practical (which includes, for example, private companies).
The personal sphere is the outermost ring, as it is the most individual and most difficult to influence; however, these spheres aren’t strictly separated and to address a particular concern or topic, sensing from various or multiple spheres are involved in the solutions.
The political sphere is in the center since it modulates between the personal and practical sphere. This is seen in the way that a city government can gather feedback from individuals about surveillance, for example, and then bring together residents and tech companies to a decision arena for negotiations around technology, laws and regulations, anonymization, data agreements, and more. The political sphere not only links the other two, but it also has the crucial function of evening weighting insights from the active (or personal) sensors and the passive (or enterprise) sensors.
Private companies that own passive sensors — and associated data — are the innermost ring as they have major leverage in changing data agreements, technology, and products. These changes have a very real impact on the ground, one of the biggest reasons that Grêt-Regamey describes them as “having so much power” and the biggest ability to steer urban transformation.
An effective sensing loop (Model B) relies on various capacities in the political sphere. One is the ability to engage residents as active sensors, which predicates a trusting and equitable relationship between cities and citizens. Another is the ability to gather sensor data from passive sensors, either through owning the technology or having a data-sharing agreement with private companies. Finally, cities must have the capability to integrate these sensors through digital platforms that facilitate the flow of information, without siloing data.
Although this sphere model can exist anywhere, this paper is specifically discussing an urban context. This is due to the fact that urbanization is “one of the most important land use change one can observe” according to Grêt-Regamey and the fact that sensing/sensors themself are passively steering how this change happens and what our urban landscapes look like. Grêt-Regamey and her colleagues have researched the process of urbanization internationally, thanks to a grant from the European Research Council (ERC); the biggest issue is the uniformity of peri-urban areas, which has led to high vulnerability of the population living in these areas across the globe with massive environmental and social problems and led to decreased sense of place and low emotional attachment to the natural world.
The authors utilized a set of diverse case studies in order to show the practical application of the conceptual model in four international cities. In Zurich, Switzerland, planners looking to develop an area of the city relied heavily on a participatory design method, despite having a strong passive sensing and mapping network. Unfortunately, the uncoupled passive and active sensing processes led to a loss of key information needed for implementation, which slowed down the process, as several intensive pilot processes were necessary to gain the support of the citizens for implementing the spatial development plan.
The flip of this situation happened in Singapore, where a wealth of passively sensed data was collected, and several active sensing campaigns were conducted to elicit and diagnose the health of Singapore’s ecosystems and the potential supply of ecosystem services as a basis to find the least environmentally impactful spot to build. While the information was integrated into a decision support platform, the sensing process ran in parallel to the traditional planning processes and only helped increase awareness for natural capital. Furthermore, the tool was fed by proprietary data belonging to the decision-makers, hindering its full use by planners.
For the authors, the best examples were cities that balanced the active and passive sensor inputs while drawing on their position as the connecting sphere between the personal and practical. In Tanzania, leaders in Dar es Salaam combined data from active and passive sensors to address flooding risks. The data from satellites, drones, and machine learning was supplemented and furthered by local flood mitigation knowledge from the local community “active sensors” which not only informed better planning but also built trust among the residents and city leaders. Similarly, urban planners in Finland city used the combination of passive and active sensing data, but took it a step further by storing both types of data into the same geospatial database. This allows multiple city departments to access layered locational information on physical data (like transportation maps or air quality) from passive sensors, and ephemeral data like children’s daily use of green space mapped by the citizen scientists.
Of course, cities must move carefully through this process in order to protect residents’ anonymity and privacy. “Cities really need to be actively involved in the production of the data themselves,” says Grêt-Regamey, who worries that “most of the cities, most of the governments don’t have that power.” The influence of the practical sphere can determine everything from what houses and areas of the city to showcase to buyers on a real estate app to which restaurant is featured in searches. This points to the need for more active involvement of the political sphere in how data is used, what kind of data is collected, and who is doing the collecting, actions that should not be solely determined by private companies. Cities can also act as a check on how diverse and representative data is, while ensuring that that data is transparently yet anonymously shared with residents through dashboards, community meetings, and open data portals.
Active sensing of the personal sphere involves eliciting citizens’ behavior and perceptions based on geospatial data collected and produced by volunteer citizens, planners, and researchers in everyday living environments. For Grêt-Regamey and her team, the most important aspect of the work is “thinking about how these sensing systems could be interrelated so that they support a values-driven planning system, and not just something driven by a few actors.” In other words, city leaders and those in the political sphere are inviting citizens to express their values when they pair active sensing data with passive sensing in decision making, a combination that produces more resilient and harmonious environments for everyone.