Our paper on ``Federated learning of predictive models from federated Electronic Health Records''...
EFRI-ARESCI: Event-Driven Sensing for Enterprise Reconfigurability and Optimization
Funding Agency: National Science Foundation (NSF), Directorate for Engineering,Emerging Frontiers in Research and Innovation (EFRI).
Award Number: NSF EFRI-0735974.
Principal Investigators: Christos Cassandras, Yannis Paschalidis and Azer Bestavros, Boston University (with R. Gao and W. Gong at UMass Amherst).
Intellectual Merit. The modern enterprise encompasses a number of processes that are all subject to changes in operating conditions, some sudden and unexpected, others slower with effects that are not immediately discernible. The proliferation of sensor and sensor network technologies provides the opportunity to enable estimators and controllers designed to rapidly react to perceived changes or detected anomalies and to appropriately reconfigure underlying enterprise components. The goal of this project is to develop a fundamental understanding of reconfigurability based on which analytical methods and explicit reconfiguration algorithms can be derived and evaluated. We ultimately envision a transformation of the enterprise towards not only flexibility (a goal that was set over two decades ago for manufacturing enterprises) but also responsiveness to unexpected, not directly observable, and possibly adversarial events. Moreover, we view the “enterprise” in the broader context of any organization whose aim it is to fulfill demand for services. In this spirit, a city is an enterprise providing services such as transportation, parking, and access to business establishments. The recent OpenAir initiative by the Mayor’s office in Boston will provide ubiquitous wireless access to the cyberinfrastructure of the internet, leading to the prospect of the “city-enterprise”. The PIs are collaborating with OpenAir and will use the wireless network deployed over Boston as one of the test beds in the project and as an opportunity to explore reconfigurability in this new type of enterprise.
Aside from the “city-enterprise” concept, the intellectual merit of the proposed project is threefold. First, it will contribute to the theoretical foundations of a reconfigurability framework for an enterprise viewed as a large-scale dynamic system. Second, it will bring together and build upon the methodological advances the PIs have made spanning: on-line performance sensitivity estimation, detection of random, adversarial and game-theoretic anomalies, robust optimization, and information acquisition systems that all capitalize on sensor and sensor network technologies. Third, this research will explore a critical shift in systems engineering with broad ramifications: replacing traditional fixed-interval, time-driven sampling and data processing by an event-driven approach better suited for large-scale asynchronous distributed environments.
The proposed multidisciplinary research is organized in three groups of tasks capitalizing on the complementary PI skills and their synergistic potential: (i) Enterprise/System tasks, for setting strategic goals and tactical performance objectives and for executing decision making processes, (ii) Network tasks, addressing communication issues in enterprise component interactions, and (iii) Sensing tasks, for information acquisition and algorithms for sensing and data manipulation.
Test Beds. An integral part of the proposed project is the application of the ideas and explicit reconfiguration algorithms developed on two complementary test beds. The first is based on collaboration with the Raymond Corp., a leading provider of forklifts to warehouses/distribution centers. We will install wireless sensor nodes on trucks in a selected warehouse with the aim of reconfiguring dispatching and repair policies in response to events such as accidents and equipment failures. The second test bed is the OpenAir wireless network over the city of Boston. We will explore reconfigurability issues in the OpenAir network through small pilot projects in parts of the city or by emulating them in an off-line setting within the Boston University (BU) Center for Information and Systems Engineering.
Broader Impact. The project will advance the state-of-the-art in all application domains that benefit from reconfigurability, primarily focusing on the enterprise where advances will result in increased energy efficiency, productivity growth, product and service quality, and enhanced workplace safety and security. We also envision a paradigm shift in the use of anevent-driven approach to data collection, setting aside outdated and inherently inefficient time-driven mechanisms; this applies to all large-scale distributed systems with asynchronously operating components. Clearly, such advances carry over to a multitude of broader areas where sensor networks and information acquisition systems are prevalent. Our proposed involvement with the OpenAir Boston initiative has the potential of revolutionizing the way we view the “city” in the future: from a passive living and working environment to a dynamic service-providing enterprise with novel applications and new ways to deal with transportation, health services, and safety. On the educational front, our plans include new courses, training graduate students, involving undergraduate students in the project, creating interactive educational software and demos, establishing cross-campus summer internships, and reaching out to high school students through two programs the PIs are involved with. Dissemination plans include capitalizing on the BUSensor Network Consortium established by the PIs, leveraging an NSF IGERT award, and organizing an academic workshop on reconfigurable systems.