ENG Transforms Curriculum for the Interdisciplinary Data Science Economy
Product developers used to depend on a primary engineering discipline to realize a design. For example, automobile manufacturers use to rely largely on mechanical engineers to design their products. More recently, companies have needed to hire a blend of software, computer, electrical mechanical and systems engineers for cars, particularly electric ones. Increasingly, the most innovative products rely both on an interdisciplinary approach and the use of massive amounts of data to support product development and operation.
Recognizing that data science is playing an increasingly central role throughout the economy — not just in cars, but in health care, urban design, the internet of things and many other fields – Boston University’s College of Engineering is transforming its approach to undergraduate education. Beginning the fall, all students will take courses that will impart foundational knowledge in data science, and they will be able to apply those tools and techniques to some of today’s cutting-edge, multidisciplinary technologies.
“The era of the single-discipline engineer is over,” said Dean Kenneth R. Lutchen. “Most innovation now requires multiple engineering disciplines interacting with large data sets. Making sure our students are literate in data analysis is fully in keeping with our mission to create Societal Engineers. I have heard from leaders in industry that data analysis is becoming a key attribute they are looking for when hiring engineers, and one not often found. Having this knowledge, Boston University engineers will have the tools to improve society for many years to come.”
In the fall of 2016, a College-wide task force was given the charge of reviewing and recommending revisions to the undergraduate curriculum in to ensure that graduates are better prepared to engage in the emerging digital and maker economy. The task force recommended changes, effective this fall, that include data science as part of every undergraduate’s education.
“We are one of the first engineering schools nationally that have designed a curriculum for which students in every major will take an interdisciplinary, data-driven approach,” Lutchen added. “We recognize it as essential and we are aware that in the future every engineering discipline, from mechanical to biomedical to computer, will intersect with data science.”
The existing two-credit Linear Algebra course will be replaced by a three-credit Computational Linear Algebra course. A single, new four-credit Probability, Statistics, and Data Science for Engineers course will replace the separate and distinct probability and statistics courses that had been offered by each department. This new course will explicitly include the tools and techniques to analyze huge data sets, and an introduction to machine learning, which is becoming a large part of autonomous systems.
In addition, three 400-level elective courses – Introduction to Machine Learning, Smart and Connected Systems, and Introduction to Robotics – will be offered. Each builds on the data science foundation and is open to students in all majors.
“Typically, engineering students are not expected to have this data science foundation,” said Associate Dean for Educational Initiatives Thomas Little. “Our new curriculum introduces them to mathematical concepts that allow them to appreciate the analytical capabilities provided by contemporary large-scale computing.”
Senior Associate Dean for Academic Programs Solomon Eisenberg added that the goal of the new curriculum is to give students a foundational understanding that will prepare them to integrate data science with engineering systems on the job or in a graduate program.
“We live in an information-rich society,” said Eisenberg, a professor of Biomedical Engineering. “We need to give students the ability to understand what is happening in this emerging area and go anywhere with it. This adds another layer of tools to the toolkit of an engineer.”
Data analytics and data-driven technological systems are playing a central role in many rapidly emerging technologies, such as smart cities and self-driving vehicles, noted Little, a professor of Electrical and Computer Engineering.
“Cyber-physical systems are producing and consuming massive amounts of data and safety is often critical. We need to be sure our graduates are prepared to use these new analytical techniques to tackle these modern challenges,” he said.
Engineers in all disciplines are dealing with increasingly large data sets, whether processing data from in-home medical sensing, developing drug molecules targeting specific microbes, designing systems that will allow swarms of robots to behave predictably, or simply making cities function better. Being able to quickly analyze the data can help take the trial-and-error out of the design process and produce products more quickly and efficiently.
In another, related curriculum change, all freshmen will design and build a product – whether physical or software – as part of the Introduction to Programming and Introduction to Engineering modules. Little said this will help build a stronger connection between the challenging first-year coursework and what students will do after graduation.
“This reflects what is going on in the maker world,” he said, “and our Engineering Product Innovation Center allows us to bring this to all students. This will transform the freshman year. Our students come to us wanting to build things and we need to keep that alive.”
The new curriculum will apply to all students next academic year. In addition to incoming freshmen, current students who have not yet taken linear algebra or probability and statistics will take the new courses that include data science.
“They need this now,” Lutchen said. “The next wave of innovations that address our societal challenges and life quality need a workforce comfortable in the era of big-data and interdisciplinarity.”