How ‘Smart Cities’ Are Building Healthy Neighborhoods

We all care about where we live. Air quality, traffic congestion, and safety are just a few of the factors that determine our quality of life. As these change over time, a ‘great neighborhood’ can lose its appeal, vitality, and residents – and the opposite happens as well. That’s why measuring urban environment, air quality, and activity patterns are a major focus of any community, not just big cities like San Francisco, Chicago, Atlanta, or New York.

Local governments have long wrestled with how to monitor and balance meeting current and future community needs effectively. Technology has helped with sensors and camera surveillance. However, the data captured could not provide deep neighborhood-level understanding of how the city and multiple factors were changing over time.

The University of Chicago and Argonne National Laboratory partnered with the City of Chicago with a vision of how to change this with Internet of Things (IoT) technology, deployed as an array across an entire city. The project, Array of Things (AoT), is a programmable urban sensor network implemented at an unprecedented scale to collect real-time data on the environment, air quality, and the flow of life in the city, from pedestrian flow to traffic congestion. With funding from the U.S. National Science Foundation (NSF), AoT was developed in collaboration with an ecosystem of citizens, government agencies, research organizations, universities, and scientists. The data is open, free of charge, for use by anyone – researchers, students, citizens, policymakers, and city departments – to investigate strategies for improving the ‘livability’ of the city.

The innovative, resilient open-source design, called the Waggle Technology Platform, is mounted on existing infrastructure like light poles. Each node contains a dozen sensors which can be remotely programmed and transmit multiple readings every minute ranging from the solar load on buildings to the density of pedestrians to flooding data, air quality, and noise levels. Artificial intelligence (AI) software running in the nodes processes sensor, microphone, and camera data to measure and analyze pedestrian and vehicle traffic, sound sources, and intensity at a level that was not feasible before. Because the nodes are remotely programmable, new AI software can be developed and added, for instance to use data from sky-facing cameras to detect drones or measure cloud cover.

AoT is an example of how IoT can provide data to help to guide policies and investments to transform ‘smart cities’ into ‘healthy neighborhoods.’ A key to operationalizing sensor data to improve the outcome of city planning, commercial construction, and neighborhood community decisions is the ability to model how the data and factors change over time. As part of the City of Chicago’s commitment to openly share data with researchers and the public, the AoT platform needed a method to analyze and share data in an accessible and usable way. The “OpenGrid” portal (OpenGrid.io) was created for public access to the City’s data as well as to AoT data.

That is where CSIRO US comes in.

After CSIRO established its Silicon Valley office, almost three years ago, the US team reached out to Argonne National Laboratory on how it could benefit from the scientific depth and breadth of CSIRO’s 5,000 scientists.

From those initial meetings, CSIRO US connected Argonne to CSIRO’s experts in Australia. That led to scientific collaboration, a CSIRO scientist on the ground in Chicago and, almost two years later, a strategic agreement. Having feet on the ground in North America made all the difference in establishing a meaningful open innovation relationship.

Part of that collaboration with Argonne resulted in incorporating CSIRO UbiSENSE into the AoT solution to enhance modeling and data analysis ability.

Developed by CSIRO Data61 with funding from the Science and Industry Endowment Fund, UbiSENSE is an analytic data platform based on physics and artificial intelligence (AI) modeling that eases the assimilation and processing of a wide range of structured and unstructured data generated from a range of sensors. The fund has been utilized to develop a Cities and Energy Challenge in CSIRO Data61 which allows researchers and engineers with a range of technical and scientific backgrounds from within CSIRO and Australian Universities to contribute to this activity.

Through UbiSENSE, data sources are calibrated then blended together and processed to deliver a better way of understanding what is happening in a neighborhood or across a city. By leveraging model-data fusion capability, it can increase the spatial and temporal resolution of AoT data streams so users can see how certain variables, such as truck traffic, change over time.

Part of the solution is Senaps, a distributed system also developed by CSIRO. It automates sensor data monitoring, capture, aggregation, and model integration, at scale. Senaps also automates model runs and provides continuous results that can be easily shared. By combining multiple datasets in a cloud environment with open APIs (application programming interface), it enables users to draw useful insights from the data.

The publish-subscribe system was designed for disparate time-series data analytics and includes an API to enable third-party developers to access and build upon the data in accordance with privacy rules^[1]. For researchers, Senaps includes a framework to deploy existing and new algorithms and models quickly.

The City of Chicago’s AoT innovation is creating interest in pilot programs in other communities around the world. CSIRO is also part of a planned program to implement AoT in Sydney, Melbourne, as well as into CSIRO-supported peri-urban and regional environmental sensoring networks in Tasmania, Victoria, and New South Wales^[2].

The potential to improve neighborhoods by learning from a global network of sensors enables communities to learn from each other on how they improve livability, resilience, and sustainability based on changing conditions.

CSIRO US facilitates relationships with US companies, government agencies, and academic institutions to connect Australian researchers with USA projects to expedite mutually beneficial opportunities for scientific advancements in food agriculture, space, water conservation, wildfire, and smart cities. Partnering in open innovation brings not only deep scientific research competencies to the table but also deep experience with a wide range of real-world problems.

[1] Senaps: A platform for integrating time-series with modelling systems, December 2017, ResearchGate
[2] Mid-Scale RI-1 (M1:IP): SAGE: A Software-Defined Sensor Network