This NSF Partnerships for Innovation: Building Innovation Capacity (PFI:BIC) project from Boston University pursued research, prototyping, and development activities to enable the rapid deployment of "smart-city" services for the city of Boston and for the Commonwealth of Massachusetts. The centerpiece of the project is a Smart-city Cloud-based Open Platform and Ecosystem (SCOPE) – a software platform for new data mechanics and analytics capabilities that support a multisided marketplace for smart-city services based on the Open Cloud eXchange model, in which potentially competing stakeholders cooperate within the same infrastructure to deliver applications and services for the public good.

The implementation of SCOPE entailed the open-source design and prototyping of new techniques and novel application programming interfaces that added capabilities that were not available in public cloud offerings. This included support for sense-and-respond real-time applications, management of data quality and provenance attributes for applications that fuse potentially noisy data from multiple trusted and untrusted sources, and the incorporation of security-enhancing services in support of privacy-preserving analytics using data from multiple stakeholders.

The project adopted a unique software development strategy by integrating the development of SCOPE into the CS curriculum through a new “Data Mechanics” course. Supervised by the PIs and by professional software engineers in the Software and Application Innovation Lab (SAIL), elements of SCOPE were developed incrementally by over 200 students who pursued over 80 projects in eight offerings of the course. Each of these projects added new components to the SCOPE platform and demonstrated its use to enable novel smart-city services and urban analytics. 

In addition to the design and prototyping of SCOPE, the investigators and their collaborators worked with partners to develop specific SCOPE-enabled smart-city services. These included (a) mobility services to reduce traffic congestion, save time and fuel, and reduce pollution, (b) energy and environmental services that monitor/estimate greenhouse gas emissions for congestion management and coordination, (c) data-driven public safety applications that leverage data assets siloed across public agencies and private companies, and (d) privacy-preserving analytics services that leverage confidential data for public good. Two notable examples of successful and on-going partnerships with the City of Boston that produced outcomes of significant societal broader impacts as a result of the use of SCOPE capabilities are summarized below.

Co-PI Hutyra and collaborators worked with the Boston Metropolitan Planning Organization (MPO) to develop a roadway-scale model of vehicle emissions of both CO₂ and five additional air pollutant species for the greater Boston metropolitan area. This model integrated EPA and Boston MPO vehicle and travel models, road sensor measurements of traffic, and a very high resolution dataset of actual vehicle travel speeds at 5-minute intervals derived from mobile phone and vehicle GPS data provided by INRIX. The final high-resolution inventory quantifies pollutant emissions at hourly timescales for every road in eastern Massachusetts. This inventory identified spatial and temporal “emissions hot spots” where combinations of vehicle activity and traffic congestion produce large amounts of pollutant emissions and also key traffic corridors where traffic congestion is responsible for localized pollution 75% above normal traffic levels. These findings were instrumental in synthesizing the “Go Boston 2030: Imagining our Transportation Future” report.

PI Bestavros and collaborators joined forces with the Boston Women’s Workforce Council (BWWC) to use SCOPE’s secure Multi-Party Computation (MPC) capabilities to support the “Boston Women’s Compact” (BWC) – a voluntary agreement in which over 200 Boston-based businesses pledged to take concrete, measurable steps to eliminate the wage gap and to report their progress anonymously every two years. The key technical challenge for the compact that the SCOPE MPC capabilities helped circumvent was to enable the calculation of pay gap analytics (at the aggregate level) without revealing the presence or extent of such gaps in any of the participating businesses. This collaboration has resulted in three iterations of privacy-preserving gender-pay-equity analytics based on real payroll information from over 120K employees – the first of its kind in the US. The aggregate data (and analysis thereof) from the 2016 and from the 2017 deployments were published by the Boston Women’s Workforce Council in 2017 and 2018, respectively. Key findings from these studies were unveiled at press conferences organized by the City of Boston Mayor (Marty Walsh) and commanded significant national media attention. The SCOPE platform component that enabled privacy-preserving gender pay-equity analytics for the City of Boston has also been used in other public-good applications by the Massachusetts Bankers Association and by the City of Boston Chamber of Commerce. SCOPE’s MPC capabilities were referenced in bi-partisan US congressional bills (including H.R. 4174 and H.R. 4479) as enablers for a host of similar social-good applications where legitimate confidentiality concerns often stand in the way of mining aggregate data assets for public good.

Last Modified: 08/12/2018
Modified by: Azer Bestavros