By Maureen Stanton
Recent UN projections show explosive growth in the urban population, doubling worldwide by 2050. It is clear that cities are on the cusp of disruptive changes. From smart phones and wearable technologies to self-driving cars, navigation apps, and drones, new smart devices that connect people, places and things are being invented every day, radically changing the ways we live, work and play.
At the center of this radical transformation is the Smart City paradigm, and it is the subject of the April 2018 special issue in Proceedings of the IEEE, the flagship journal of the Institute of Electrical and Electronic Engineers. This special issue presents recent advances and technical solutions geared toward implementation of Smart Cities and includes two papers authored by several affiliated faculty and students of the Center of Information & Systems Engineering (CISE) at Boston University College of Engineering.
“Cities are looking for ways that ensure a sustainable, comfortable, and economically viable future for their citizens by becoming ‘smart,’” said Christos Cassandras, one of the three invited guest editors of the Proceedings of the IEEE Smart City special issue, Head of the Division of Systems Engineering (SE), and Professor of Electrical and Computer Engineering (ECE) at Boston University.
With the rise of urbanization comes traffic congestion, making transportation a key area of focus. Professor Cassandras led one of the first groups in the country to design and model a Smart City and has continued to be at the forefront of developing new technologies to allow efficient, safe, smart transportation for cities of the future.
Cassandras, along with CISE Director Ioannis (Yannis) Paschalidis (ECE, SE) and former CISE students Jing Zhang (Ph.D. 2017) and Sepideh Pourazarm (Ph.D. 2017) are authors of a paper entitled “The Price of Anarchy in Transportation Networks: Data-Driven Evaluation and Reduction Strategies” that appears in the Proceedings of the IEEE special issue. This new paper demonstrates that during heavily congested traffic periods, the use of socially optimal routing can lead to as much as a 50% reduction in congestion. The researchers used minute-by-minute traffic data from the Boston area to estimate the effect on traffic congestion of drivers’ selfish route selection as opposed to a more coordinated, socially optimal routing scheme. (Read more.)
As the world rapidly urbanizes, there is increasing focus on the potential health risks of city living. “City living and its increased pressures of mass marketing, availability of unhealthy food choices and accessibility to automation and transport all have an effect on lifestyle that directly affect health,” according to The World Health Organization, which is calling for a shared effort to put health at the heart of urban policy.
The Proceedings of the IEEE Smart Cities special issue included several papers that demonstrate how advanced engineering, and information and communication technology (ICT) can address community health problems, including a new paper entitled “Predicting Chronic Disease Hospitalizations from Electronic Health Records: An Interpretable Classification Approach” authored by Professor Paschalidis and William G. Adams, M.D., Boston Medical Center and Director of BU-CTSI Clinical Research Informatics for Boston University, along with former CISE students Theodora Brisimi (ECE, Ph.D. 2017) and Wuyang Dai (Ph.D. 2015), and current CISE students Tingting Xu (SE) and Taiyao Wang (SE). The study focuses on the two leading clusters of chronic disease, heart disease and diabetes, and develops data-driven methods to predict hospitalizations due to these conditions.
In this study, the researchers validated the algorithms on large data sets working with over 50,000 Electronic Health Records (EHRs) from the Boston Medical Center, the largest safety-net hospital system in New England. They found that they could predict hospitalizations from these two chronic diseases about a year in advance with an accuracy rate of as much as 82% —a marked improvement over a predictive model using the Framingham Risk Score (56%), the gold standard for predicting the likelihood of heart disease.
“The potential benefits from applying machine-learning analytics in health care are enormous,” said Professor Paschalidis. “By giving care providers the chance to intervene earlier and head off hospitalizations, quality of life can be improved significantly. Preventing hospitalizations in cases of these two widespread chronic illnesses alone — heart disease and diabetes — the United States could save billions of dollars a year.”
Professor Paschalidis is involved in a number of e-health research projects, including collaborations with Boston Medical Center and Brigham and Women’s Hospital. In addition to preventing hospitalizations, Paschalidis and his research teams have applied machine learning to health care challenges such as predicting and preventing readmissions after surgery, automatically controlling medication dosage at the ICU, and identifying CT scans where the patient received more radiation than what was medically necessary. Professor Paschalidis includes highlights of some of this recent work in his Harvard Business Review article entitled “How Machine Learning Is Helping Us Predict Heart Disease.”
Collaborating with experts in academia, government, and industry, CISE affiliated faculty and students are advancing Smart Cities systems and technologies as well as exploring economic, environmental, and public policy implications. Learn more by visiting the CISE website.
The Smart City special issue appears in Proceedings of the IEEE Volume 106, Issue 4 | April 2018. The table of contents with links to full article abstracts can be accessed here.