|
MIT Portugal Program research in Engineering Systems provides a mechanism to explore interdependencies and extend the methodological approaches across our areas of research. We examine interdisciplinary approaches and methodological areas, as well as new methods and research tools beyond those currently used in engineering.
As part of the Program’s revisited strategy, Fundamentals of Engineering Systems was defined as one of our targeted application areas. Click here to learn more about our research program in this area.
Earlier MIT-Portugal research in fundamentals of Engineering Systems included projects in the following areas:
- System and Design Architecture
- Enterprise Perspectives
- Uncertainty/Flexibility in Design, Deployment and Operation of Engineering Systems
- Networked Infrastructure Systems
- Supply Chain Efficiency
1. Networks
Distributed Inference in Critical Infrastructures
This is a joint project between MIT and CMU focusing on the use of sensors in infrastructure systems. Co-PIs: Sanjoy Mitter (MIT); Jose Moira (CMU).
This project considers the problem of distributed inference and actuation in large-scale critical infrastructures. It emphasizes energy systems, with project team researchers at MIT, CMI, and ICTI in Portugal. Critical infrastructures, like the power grid, are complex dynamic networks evolving in real-time, where events at any location are often correlated with other events in space/time. We are concerned with the problem of inferring the general global state of the critical infrastructure from the distributed measurements collected by a network of sensors. The main challenge can be summarized as follows: How to make decisions under uncertainty arising from spatially- distributed dynamic information when sharing distributed data is limited by networking constraints. The research focuses on developing mathematically rigorous techniques for reliable monitoring of events, developing situational awareness, and providing prediction of abnormal behavior like threats and failures.
Network Design & Optimization
PI: Thomas Magnanti (MIT) along Luis Gouveia (University of Lisbon)
This project would develop optimization approaches (models and solution procedures) for large-scale complex network systems that arise in transportation or telecommunications settings. It would focus on problems that are combinatorially complex such as those that occur in vehicle routing and scheduling, or in the design of cost efficient and reliable telecommunications systems, such as the Internet. The project would focus on generic models and solution approaches that could be applied in many specific application settings. The research would build upon collaboration of several years by the two co-PI that have addressed issues in these general problem domains.
Networked Systems Engineering Systems Modeling and Visualization for Networked Systems
Profs. Olivier de Weck and John Hansman
Critical networked infrastructures such as air and ground transportation, communications, power, financial systems, water, and health care are example of complex socio-technical systems. At an abstract level many of these systems exhibit similarities in their structure evolution, operating dynamics and vulnerabilities. There has been significant research on specific infrastructure problems but for some time the MIT ESD has been interested in a more generalized comparative approach to critical network infrastructure where insights from one network domain is used to stimulate insight and opportunities in other network domains.The objective during Year 2 will be to identify key network infrastructure problems in Portugal. Follow-on work will then focus on the development of an integrated modeling and visualization capability for complex networked systems. This work would directly support research in the four focus areas as well as be one of the mechanisms by which linkages are created between the four areas. One thing that all four areas have in common is that they involve network representations of systems. Research will focus on applying network modeling algorithm and tools including network modeling and analysis; agent-based simulation, discrete event systems simulation; and data mining and validation.
2. System Design and Architecture
Application of Dynamic Multi-Attribute Tradespace Exploration to the Architecting and Design of a Transportation Engineering System
Dr. Donna Rhodes and Dr. Adam Ross
The engineering of complex systems involves sophisticated decision analysis under conditions of high uncertainty, requiring many variables to be taken into account. Some systems (e.g., manufacturing) lend themselves to single-decider approaches, while others (e.g., the transportation system) have multiple decision makers. A need exists for a holistic Engineering System design process to take into account changing system contexts and value expectations, allowing for the quantitative comparison of many possible design options during the concept exploration phase of design. This proposal seeks support to apply the Dynamic Multi-Attribute Tradespace Exploration (MATE) method to the architecting and design of an Engineering System in the Transportation focus area. Past MATE studies have been conducted in the aerospace domain and it is anticipated that application to a new domain area will uncover new issues that face Engineering System architecture and design, but may not be apparent within the context of aerospace systems.
3. Supply Chain
Carbon-Efficient Supply Chains
PI: Prof. Yossi Sheffi (MIT)
The goal of this research is to develop a methodology for characterizing the carbon intensity of supply chains. Specifically, we are interested in addressing the following questions:
- Which are the key carbon metrics required to describe a supply chain? How should they be calculated (e.g. allocation)?
- What is the effect of “supply chain scope” when determining carbon efficiency?
- What is the effect of time-horizons on supply chain carbon metrics and scope?
- How do carbon metrics affect traditional supply chain decisions (e.g. outsourcing, off-shoring, and transportation mode)?
- How do carbon-efficient strategies vary by supply chain characteristics?
- What are the key elements and challenges in creating a “carbon label”?
- How can you build a “business case” around carbon-efficiency initiatives on supply chains?
The project will continue refining the carbon measurement methodology. A new series of case studies will be developed with special emphasis on scorecards and incentives.
4. Enterprise Perspectives
Building Lean Enterprise Educational Curricula
PI: Prof. Deborah Nightingale
In line with the Portuguese government’s intention to strengthen the country’s knowledge base at an international level, it is crucial to create and teach educational materials that promote the planning, development and implementation of “lean enterprises.” In lean enterprises, the dimensions of people, technology, processes, and information are addressed in unified framework, with emphasis placed on the integration of these dimensions across the enterprise (product development, production, supply chain, etc.). MIT has introduced the analysis and implementation tools into the teaching materials that have been developed in this area. During Year 2, this project will assemble the various teaching materials that have been used in different courses at MIT and augment them with new case studies that may include lean manufacturing, lean engineering, and knowledge management. Case studies will make use of examples from Portuguese industry. A textbook will ultimately be produced.
5. Uncertainty/Flexibility
Text Book on Maximizing Value Through The Use of Flexibility in theDesign Development and Operation of Engineering Systems
Prof. Richard de Neufville
Traditionally, engineering practice designs systems to specifications set externally by a commercial or political agenda. There is typically an implicit assumption in these designs that specified functions or loads are static over time, and so they generally fail to provide the means to transition to new requirements effectively. The new approach recognizes that “forecasts are always wrong,” and emphasizes design for flexibility so that system can be modified either to exit from undesirable states, or to take advantage of new opportunities. The proper use of flexibility provides major increases in expected value (around + 20% in documented cases).The ultimate objective of this project is to produce a new textbook on flexibility in systems design. Specific work to be initiated during Year 2 includes developing analytical procedures for assessing flexibility; developing methodologies to delineate and assign value to alternative forms of flexibility; and developing analytic and design procedures to incorporate flexibility objectives into design processes. Prof. de Neufville plans to spend sabbatical time in Portugal in order to develop Portuguese-specific examples and cases to be used in the research.
Education Models, Data and Inference in Socio-Technical Systems
Profs. Richard Larson and Daniel Frey
This is a curriculum project to develop a course on quantitative methods for engineering systems. This will become a required part of the core curriculum for all ESD PhD students at MIT, and we will work to incorporate it, all or in part, into the educational programs of the different focus areas. The course will cover methodologies and concepts such as Poisson and Markov processes, applied probability and statistics, hypothesis tests, regression, correlation and causation, simple data mining techniques, and Bayesian vs. classical statistics.
Social Science Research Methods for use in Engineering Systems Research
Prof. David Mindell
This is also a curriculum development project that will produce modules aimed at introducing social science research methods to engineers. Like the quantitative methods course, it will be used for ESD doctoral students and we will work with the focus areas to incorporate it into Portuguese educational programs. EDAM has already expressed interested in using this course and it is expected to be of interest to others as well.
|
