Since 2011, NSF’s Innovation Corps has sent 365 research teams out of their labs to scout potential customers, and helped launch 208 start-ups.
Pig poop is not a substance most engineers typically deal with. But during a four-year period beginning in 2008, Elham “Ellie” Fini, an assistant professor of civil engineering at North Carolina A&T State University, figured out how to put porcine manure to practical use instead of shunting it to waste-treatment plants. She developed a thermochemical liquefaction process that turns it into an adhesive, while simultaneously sequestering any carbon dioxide and other greenhouse gases created during manufacturing. It’s a clean, green product that’s also 75 percent less expensive to make than petroleum-based glues. And given that American pigs produce around 6 billion gallons of waste a year, there’s a ready supply of feedstock.
Fini initially thought the only market for PiGrid, as it’s called, would be as a binder in manufacturing asphalt. But in 2012, Fini formed and led a three-person team with a remit to talk to around 100 potential PiGrid customers. That’s when her team — which included Mahour Mellat-Parast, an assistant professor of technology management — quickly discovered that asphalt manufacturing was only one of many potential markets for PiGrid. Other uses included bookbinding, packaging, and flooring. “It was an eye-opening experience,” Parast recalls. The team has since formed a company, Bio-Adhesive Alliance, that’s received more than $500,000 in grant and award money and expects to open a pilot plant by mid-2015.
Getting out of the lab and talking to scores of putative customers proved to be a savvy business move, but it’s not a common practice among researchers. Fini’s team took the plunge after joining the National Science Foundation’s Innovation Corps (I-Corps), a program created to move inventions and products derived from NSF-funded basic research to the marketplace. In I-Corps, researchers find out if the results of their work have commercial potential and learn the basics of starting a company.
Launched in 2011, I-Corps was a signature project of Subra Suresh, then director of NSF and now president of Carnegie Mellon University. Suresh had the notion of pushing principal investigators “to get them to a place of higher commercial visibility,” according to Tom Peterson, who led NSF’s Engineering Directorate at the time. This was considered by some a radical departure from the agency’s primary mission of funding basic research, says Peterson, now provost and executive vice chancellor at the University of California, Merced. While NSF-backed research is credited with such world-beating technologies as bar codes, Internet search engines, and magnetic resonance imaging, commercially promising findings often never leave university labs. As Suresh noted in 2011, the process of capitalizing on discoveries “eludes us more than is necessary or desirable.” Not specifically authorized by Congress, I-Corps was funded at a modest $50,000 a year for each of 100 researchers. This represented only around a quarter of 1 percent of the NSF’s budget, Peterson says, hardly detracting from NSF’s “bread and butter” role of fundamental research.
A Silicon Valley Guru
“One of the major goals of the program is increasing the economic impact of NSF-funded research,” says Rathindra DasGupta, I-Corps’s program director. The best way to do that, the NSF reckoned, was to prod academic researchers — who, by and large, are often clueless about how businesses and markets operate — into thinking more commercially. But helping to create small start-ups from scratch isn’t among the NSF’s skill sets. So NSF enlisted as partners the Kauffman and Deshpande Foundations, both of which encourage innovation and entrepreneurships as part of their mission.
To provide training, NSF turned to Steve Blank, a successful Silicon Valley entrepreneur who has become a leading business guru, and asked him to create the I-Corps curriculum. Blank has had his hand in eight different start-ups, both hits and misses. He hit a homer when he co-founded Epiphany, a business software company that sold for $329 million in 2005. Since then, he’s gone on to write several how-to books on entrepreneurship, and teaches at both Stanford University and the University of California, Berkeley.
Blank modified and streamlined his Lean Launchpad course for I-Corps so it would better fit the busy schedules of PIs. The heart of Blank’s system is sending researchers into the field to speak with potential customers – lots of them. The point is not so much to promote their technologies but to listen and learn what those customers’ needs and problems are, and then, if possible, find a way to use their inventions to meet those needs. They also learn to quickly change or drop an idea and move on if they realize they can’t give the customers what they want. “A lot of people think entrepreneurialism is an execution process, but it’s not,” Blank says. “It’s a search process, a search for a business model.”
Any academic who has received research money from NSF within the past five years is qualified to apply to I-Corps, which last year had a budget of $12.5 million. Each quarter, up to 25 teams are selected, and each is awarded a $50,000 grant. PI applicants have to assemble a team comprising themselves, a graduate student who is designated the “entrepreneurial lead,” and a mentor who is a businessperson local to them. Student involvement is key because if the team opts to start a company, it’s usually the student who will run it. “Ninety-nine percent of these PIs will say, ‘Look I don’t want to leave my day job’” in academia, even if they want to commercialize their technology, DasGupta says. The screening process is tough. Applicants are first interviewed by DasGupta and other NSF program directors, and then by I-Corps inspectors. Interviewers look for teams whose basic research is completely finished and who have a real passion to succeed. Accordingly, only a relatively small percentage of applicants are chosen. Says Blank: “By the time they get to class, they are self-selected. They really, really want to be in this class.” This is important, because the the seven-week course is grueling and time-consuming, and requires a huge commitment. “We push them hard,” Blank says.
Boot Camp and Cold Calls
Initially, Blank and a team he put together taught the course from Stanford. It’s now taught both in class and online from one of seven “nodes,” which include the University of Michigan and Georgia Tech. Stanford now is part of a Bay Area node that also includes Berkeley and the University of California, San Francisco. There are also regional nodes in the New York and Washington, D.C., areas, as well as in Southern California and the Southwest (mainly Texas).
The course opens with a three-day “boot camp” at one of the nodes. That’s a good way to teach entrepreneurship, says Robert Langer of MIT. “The best way to learn is by doing.” Langer ought to know. The professor of biochemical engineering has been involved in 26 startups, whose products range from medical diagnostic tools to hair-care products.
At boot camp, the academics and their teammates are taught how to talk to customers, including how to make eye contact. Some PIs bristle first at the notion of getting out of the classroom and cold-calling customers, Blank relates, and say they’ll send their grad student out to make all the calls. “I say, ‘No f… way’”—the PIs have to make many of those calls themselves because no one understands the science and technology behind the core idea better than they do. As Elham Fini can attest, these conversations can make a crucial difference in how a researcher views the marketplace. Indeed, it is the rare PI whose initial product idea is still intact at the end of seven weeks. It helps that academics are accustomed to testing a hypothesis, learning from failure, pivoting to a new hypothesis, and testing anew.
Teams also get a crash course in business. Among the things they learn: what a business model is and how to test it; who their customers are; how to design a distribution channel; how to create demand; what’s the revenue model; and whether partners are needed. “No one understands this stuff going in. Why would they?” Blank says. “But we’ve cracked the code on how to teach entrepreneurialism to scientists and engineers without wasting their time.” At the end of the course, the teams themselves decide whether to move ahead. A “go” doesn’t necessarily mean starting a company and seeking an NSF Small Business Innovation Research (SBIR) grant or other investments; other viable options include partnering with a larger company or striking a licensing deal.
I-Corps has so far put 365 teams — nearly all led by NSF-funded PIs — through boot camp, and around 208 of those teams have gone on to create start-up companies. To the extent that many of these teams might otherwise not have sought commercialization, these numbers count as a success. But it’s still too soon to know how many will turn a profit. Of 130 teams formed in the 2011-12 period, about 20 reported that they had received some sort of equity financing — from venture-capital companies or other nonfamily sources — and had raised a total of $9.2 million.
Without I-Corps training, DasGupta says, too many researchers launch companies based on products that have no market, all the while burning through thousands of dollars or more of financing. “We are saving PIs tons of effort and money,” Blank says. He told a congressional committee in 2012 that I-Corps “is an extraordinarily efficient use of taxpayers’ money. It will pay us back with jobs and a competitive edge on a global scale.”
Some teams apply for SBIR grants, for which their success rate is higher than the 18 percent average, NSF says. This was the path pursued by Jizhong Xiao, a professor of electrical engineering at the City College of New York (CUNY), who invented a wall-walking robot. He says the I-Corps class was worth the sweat. “It was exactly what we needed, and it forced us to get out of the building,” Xiao says. “Without it, I believe this technology would still be on the shelf.” Xiao first wanted to commercialize his robot back in 2008 — that’s when he unsuccessfully applied for an SBIR. He thought the market for his robots was owners of tall buildings — those over six stories — that in New York, by law, have to be inspected every five years. Why not let robots do it?
Xiao dusted off his business plan when he successfully applied to I-Corps. After hitting the streets to talk to customers, his team discovered that building owners didn’t care how the inspections were done. But the companies that landlords contract to carry out inspections were intrigued by the idea of using lightweight robots that could do the job for much less money than humans. Team Xiao further discovered that the inspection market was bigger – much bigger – than buildings. It also included a wide variety of infrastructure, including dams, bridges, tunnels, pipelines, and power plants. Accordingly, Xiao launched a start-up, InnovBot, which has so far gotten a Phase I SBIR grant to improve the robots’ performance. “Some customers are so excited that they want to put in conditional orders. That’s a good sign,” says Xiao, who expects to apply for a Phase II grant.
Model For Other Agencies
The I-Corps model has now been adopted by both the National Institutes of Health and the Department of Energy. An NIH team sent through the NSF program earlier has already scored some success. Led by researchers at Carnegie Mellon, the team went on to start Neon Labs, which uses algorithms based on cognitive neuroscience that autonomously figure out which frames from a video will most appeal to viewers. The results are turned into thumbnail pictures and used to drive traffic to online video publishers. Earlier this year, Neon Labs received its first major round of venture capital: $4.1 million. This fall, DOE announced it was launching Lab-Corps, which initially will give priority to clean-energy technologies with teams selected from its national laboratories.
NSF is trying to scale up I-Corps’s training without having to expand the basic national program. Already around 70 percent of I-Corps participants make an effort to pass on what they learned to their colleagues and students. To boost and standardize those efforts, the agency has contracted VentureWell to run logistics and training management. Formerly the National Collegiate Inventors and Innovators Alliance, begun in 1995, VentureWell provides grants to faculty to teach technology entrepreneurship. “I-Corps participants have an appreciation for this stuff and want to share it. That’s having a significant impact,” says its president, Phil Weilerstein. VentureWell recently worked with Blank to develop a faculty training program at the nodes. NSF has 16 I-Corps sites across the country, ranging from the University of Toledo to the University of Utah, which receive mini-grants to encourage early-stage innovators — faculty and students — often by using scaled-down versions of Lean Launchpad. VentureWell also coordinates a National Innovation Network that connects the sites and nodes and makes the sharing of knowledge and training ideas easier.
What about the PIs who decide it’s a no-go? “They’re still great scientists,” Blank says. Despite their disappointment in finding their idea wouldn’t work, “they’re happy about it because they didn’t waste a lot of time and money finding out.” Moreover, once they’ve been schooled in the Blank method, they’ve gained a skill for life that they can apply to future research.
Consider Sunny Shah, a biomedical engineer. Shah was a senior scientist at Notre Dame University’s Advanced Diagnostics and Therapeutics Initiative, a program that develops health care technologies, when in January 2013 he joined a three-member team — led by Hsueh-Chia Chang, a professor of chemical and biomolecular engineering — to determine if there was a market for a detection system he had invented with Chang that quickly sniffs out the DNA and RNA of dangerous pathogens lurking in foods. Shah spent many hours outside his lab, interviewing scores of butchers and food processors. “I was completely out of my comfort zone,” he admits, although he eventually found he enjoyed the process.
But ultimately his team was one of only two in its cohort that opted not to proceed. While the technology it offered could detect pathogens more cheaply than current methods, what customers wanted was
something faster, not cheaper. Shah admits it was “heartbreaking” to reach the no-go decision, but it saved the team from throwing good money after bad on a product no one wanted. Shah and Chang are looking to see if there are medical markets for the technology. One possibility: using it for early detection of oral cancers. Another possibility: detection of dengue fever in developing nations. The customer-centric approach, Shah says, “is a process I pretty much live by now.”
Shah, currently assistant director of Notre Dame’s Engineering, Science, and Technology Entrepreneurship Excellence Master’s Program (ESTEEM), has developed and teaches a course based on his I-Corps training. He continues to spend around 20 percent of his time on research, but with an eye open to possible products. “I wouldn’t have thought that way two years ago,” Shah confesses. That’s one other lesson academics learn from I-Corps. Once bitten by the entrepreneurship bug, you can’t shake it off.
By Thomas K. Grose
Thomas K. Grose is Prism’s chief correspondent, based in the U.K.
Design by Nicola Nittoli
Illustration courtesy of thinkstock