Creating an Engine for Breakthrough Innovation in STEM Education

Every organization can benefit from an internal group that focuses on promoting and creating game-changing innovations.1 To avoid falling behind, organizations must look to the future while also improving performance and practices in the present. Here at the U.S. Department of Education (ED), we’re working hard to build the foundation for an advanced research infrastructure that can uncover breakthrough innovations so that our schools, educators, and students once again lead the world.

Before joining the team at ED, I spent 22 years in different Department of Defense (DoD) research settings, working closely with a variety of civilian research agencies. What I learned leading projects at the Defense Advanced Research Projects Agency (DARPA) and the Office of Naval Research (ONR) is that most research (both public and private) is stove-piped into two categories: basic and applied. Basic research seeks new knowledge and understanding, while applied research — as the name suggests — takes existing knowledge (i.e., the results of basic research) and creates new applications for it. Applied research can improve performance incrementally by leveraging the results of already-established basic research. This is an important and essential function. But by definition, the impact of applied research is limited by the horizon of current knowledge, which means it is not well-suited to producing dramatic breakthroughs.

BreakthroughIn the education sector, unfortunately, we underinvest in both kinds of research at every level — across the public, private, and nonprofit sectors. (Compared to our peers in healthcare or energy, we spend just fractions of our resources on research and development.) And while American educational achievement and attainment is improving, our progress is far too slow. To regain our leadership in high school and college graduation rates, we need new research and development tools that will lead to breakthrough innovations that dramatically improve student outcomes and tools that go to scale to serve students across the nation.

How do we maximize opportunities to create breakthrough innovation? There is a model, which has been highly successful for more than 50 years, but has not been widely replicated in any sector: DARPA. Immediately prior to joining ED, I was a program officer at DARPA developing and managing programs on psychological health, training, and education. DARPA was created in 1958 in response to the Soviet launch of Sputnik. Its mission is simple, but powerful: to prevent technological surprise by being an organization that itself creates technological surprise. DARPA programs have created numerous successes that have dramatically enhanced military might and effectiveness (such as stealth technology and night vision), while other programs led to breakthroughs such as the Internet and GPS that have fundamentally reshaped society.

The DARPA model could similarly boost education research, especially in the context of science, technology, engineering and math (STEM) education. Just as we created DARPA to keep the United States at the forefront of technological advancement, we must pursue advanced education research projects to create breakthrough innovations to ensure that future generations of Americans have the skills and abilities they need to compete in and lead the world. Advancements emerging from this process would create the next generation of innovative, highly trained scientists and engineers to sustain a significant technological lead. It would also help to create an education system that promotes lifelong learning to enable U.S. workers to continue to adapt to rapidly changing technology environments and remain competitive.

How is the process used at DARPA different from the more traditional basic and applied research? DARPA employs a research process that resides in Pasteur’s Quadrant, a category first described by Donald Stokes.2 Stokes broke research into four categories that balance the search for basic understanding with the aim of producing a specific outcome. Basic research, classified as Bohr’s Quadrant, is the quest for basic knowledge without regard for the final use of that knowledge. Applied research resides in Edison’s Quadrant, where producing a specific product is the top priority; effort is dedicated to “applying” existing basic research. Pasteur’s Quadrant collapses the boundaries between Bohr and Edison: it conducts basic research aimed at solving specific and immediate problems. In that sense, basic research in Pasteur’s Quadrant is driven in real-time to meet the ongoing requirements of the final application.

Pasteur's Quadrant fills the gap between Bohr's quest for basic knowledge without a final use in mind and Edison's emphasis on specific products, offering the possibility of breakthrough innovations. (Photo credits: Bohr and Pasteur photos courtesy of the Smithsonian Institution; Edison photo courtesy of the National Park Service)

Pasteur’s Quadrant fills the gap between Bohr’s quest for basic knowledge without a final use in mind and Edison’s emphasis on specific products, offering the possibility of breakthrough innovations. (Photo credits: Bohr and Pasteur photos courtesy of the Smithsonian Institution; Edison photo courtesy of the National Park Service)

Pasteur’s Quadrant is interesting because it’s based on having a clear vision of success even when all of the pieces of the puzzle are not yet known. As a result, success often means challenging prevailing assumptions and leapfrogging over current practice to produce a breakthrough. The race to the moon was very much a process of breakthrough innovation. The U.S. set an audacious goal whose achievement was beyond the capacity of the science and technology of the day. Basic and applied research therefore needed to iterate and develop continuously to achieve the goal. Much has been written about the race to the moon, but what is clear is that if these two forms of research had been separated, we never would have achieved our goal.

Projects created within a DARPA model — like those necessary to win the race to the moon — do not fit well within traditional research management structures in which basic and applied research are separated. Typical applied research programs require specific milestones and clearly defined deliverables. Project details remain fairly static over the course of a project or program. In the DARPA model, by contrast, every project is a mini-moonshot. The final goal is clear, but the process for getting there remains nimble to account for what is learned during the research process and what new challenges may arise.

How do we use this process to create innovation in education? We bring together interdisciplinary teams of world-class experts with proven track records of innovative thought and action. It requires a balance of expertise, flexibility, discipline, collaboration, and creativity along with a visionary program officer to lead the work of these experts according to a rigorous program plan. Performers are given plenty of room to be creative while progressing toward the established goal.

How could the U.S. Department of Education accelerate and spur breakthrough innovation? With the help of top experts and leaders in the field, ED’s new STEM office is developing a vision of what a reimagined approach to STEM education might look like in 2025. Metrics of success will be developed to drive progress towards that vision. From this process, we will identify key hurdles that need to be overcome to achieve that vision, with a particular focus on how technology could be catalytic. Working within that frame, advanced research projects for education could be used to achieve key goals within a three-year time frame tied to the metrics of success. The same criteria used to create DARPA programs would be used to identify such programs. Known as Heilmeier’s Catechism after a former DARPA director, the criteria are simple and straightforward:

  • What are you trying to do? Articulate your objectives using absolutely no jargon.
  • How is it done today, and what are the limits of current practice?
  • What’s new in your approach and why do you think it will be successful?
  • Who cares?
  • If you’re successful, what difference will it make?
  • What are the risks and the payoffs?
  • How much will it cost?
  • How long will it take?
  • What are the mid-term and final “exams” to check for success?

For example, one “education moonshot” might be to develop education technology that enhances and transcends the boundaries of current classrooms — improving the classroom experience but also assisting with lifelong learning objectives as our life and career trajectories develop. Imagine if we each possessed lifelong learning tools that are as ubiquitous as the Internet or GPS. And imagine that these education tools merged with decision-support tools to help us in our careers. Teachers, engineers, doctors, and technicians would all have tools to help keep their skills fresh and up-to-date.

Breakthrough 2It’s an exciting vision. But what are the building-blocks — the “mini-moonshots” — that must be in place? What hurdles must we overcome now to achieve that vision? One example might be improving “real-time” embedded assessment technology that measures practical knowledge on demand and enhances data-driven instruction. In the classroom context, this could eliminate the need for standardized testing as we know it. Another example might be creating personalized, intelligent tutors using highly customized approaches that are tailored to instructional needs of specific students and that adapt tutoring across entire curricula. Both of these examples would simultaneously improve classroom instructional capabilities and expand the horizon of what is possible in the classroom and beyond. Like DARPA, education research should be aimed at solving complex problems. Like DARPA, even projects that fall short of their ultimate goals would move the needle in key research and development areas that can have profound impact.

Why the Department of Education? DARPA and other federal agencies are leading groundbreaking innovations by building an array of models for advanced research. At the Department of Energy, an ARPA for Energy (ARPA-E) is developing the next generation of battery technology and electrical grids that will power the American households and businesses of the future. At the U.S. Agency for International Development, the Global Development Lab is seeding new solutions to end global hunger, eradicate disease, and build stronger communities across the world.

Each of these efforts shares a commitment to bringing together diverse experts from the private sector, universities, government, and nonprofit partners to identify what works and bring it to scale for a fraction of the time or cost of traditional government programs. Still, research in Pasteur’s Quadrant is the exception in the public and private sectors, and is completely absent in the education sector. The Federal Government can and must lead by example, tackling the toughest challenges and developing solutions quickly and efficiently to produce quantum leaps in educational practice.

Breakthrough 3Taking on this work requires marshaling resources in pursuit of ambitious goals, the flexibility to develop new ideas around failed ones, and most important, the capacity to leverage crosscutting expertise in the field. ED has unique access to the nation’s top experts in research settings and educators that are championing innovative approaches for their students every day. And, if ED was equipped to pursue advanced research projects, it would have the mandate, capacity, incentives, and resources necessary to lead innovative research at scale.

Leadership in American education research and development is just as essential for national security and prosperity as military success. Just as DARPA was created to ensure that the U.S. was not caught by technological surprise, success in 21st century education requires the adoption of new approaches that can drive groundbreaking research to equip students, teachers, and families with the tools they need to succeed in a dynamic STEM economy that requires all citizens to be lifelong learners.

Russell Shilling is executive director of STEM in the Office of Innovation and Improvement.


1Past leaders of the Defense Advanced Research Projects Agency (DARPA) have emphasized this point. Dugan, Regina E., and Kaigham J. Gabriel. Special Forces’ Innovation: How DARPA Attacks Problems. Harvard Business Review 91, no. 10 (October 2013), pp. 74-84, available at

2Stokes, Donald. Basic Science and Technological Innovation. Brookings Institute Press (August 1997), available at