While many big companies still utter the words “open innovation” with trepidation, getting bogged down in discussions about who will own what, General Electric has been racing ahead. Few companies are amassing more experience in leveraging the wisdom of the crowd for novel solutions to manufacturing problems.
In 2013, the company took to the Internet and challenged engineers around the world to improve on its design for brackets that are affixed to jet engines. To get their shot at a $20,000 prize pool, competitors had to design a bracket that met GE’s strength and shape requirements and could be crafted through additive manufacturing. That is, rather than being carved from a block of solid metal, the parts would be 3D-printed from a titanium alloy powder, melted into shape with a powerful laser.
The top prizes would go to the brackets that weighed the least, compared to GE’s existing 4.48 pound brackets.
“If you are familiar with the aviation industry, weight equals fuel consumption equals cost to our aviation customers,” explains Dyan Finkhousen, who heads GE’s Open Innovation Center of Excellence.
The contest received more than 600 entries, with the grand prize going to M. Arie Kurniawan, an engineer from Indonesia with no prior aviation design experience. His bracket design weighed just 0.72 pounds, cutting the part’s mass by nearly 85 percent.
“For us, the results were really a seminal moment,” says Finkhousen. “We saw the sheer power of just opening up in a very constructive way some of our business opportunities, and really collaborating broadly with these global expert communities.”
Taking the Next Step
The company saw similar results from contests in other fields. One challenge, dubbed Flight Quest, focused on a type of aviation problem more familiar to the flying public: predicting flight arrival times more accurately. GE received more than 3,000 submissions, as teams wrote code to crunch the numbers and compare their predictions to the competition and to industry standards.
“Then they would go back and train their models, refine, and re-upload,” says Finkhousen. “The winner of that particular competition beat the industry benchmark by 34 percent, and had no aviation experience.”
The top entries were incorporated by GE’s global research and innovation teams, which are now working to produce a commercial product, she says.
The Open Innovation Team at GE
Finkhousen runs a three-person team at GE’s Fairfield, Conn., headquarters that helps other groups within the 300,000-employee company run those types of open challenges to crowdsource solutions.
“Right now, the team is really comprised of program managers — folks who are very strong in their consulting skills,” she says. The company plans to expand the team to eight members by the end of 2015 and continue to hire next year, bringing on experts in engineering, materials sourcing, and product design.
Here’s how Finkhousen defines open innovation:
A ‘Translation Layer’
Finkhousen’s group forms a “translation layer” between other GE teams and outside specialized vendors that provide platforms for open innovation competitions, like GrabCAD for computer-aided design tasks (like the bracket challenge), and Kaggle, a data science network that worked with GE on the flight arrival competition. Other vendors have included crowdsourced research firm 10EQS, collaborative hardware design company Local Motors, and open innovation consultancy BigHeads Network.
“To encourage adoption of the new methodology, we’ve built a consistent, compliant open innovation process that we use to provide these services to all the GE teams,” she says. “We’ve designed our open innovation process into a standard open innovation agreement that we can use with all our vendors.”
That process means first working within GE to figure out the precise goal of each challenge and other parameters — like whether the campaign will be open to the public or by invitation only, whether GE will be disclosed as the sponsoring organization, and who will get the rights to the intellectual property produced.
“In some cases, we may want to acquire ownership of the intellectual property, and in other cases we may just want to be on a discovery journey, and of course there’s always a range in between,” says Finkhousen.
One recent challenge focused on helping GE determine its options for sourcing a particular class of advanced materials that are used in healthcare imaging. Finkhousen isn’t at liberty to discuss the project in too much detail, but she says the challenge spurred 22 vendors to provide proposals to supply the product, for which GE initially had only one likely supplier.
GE then gave the top 10 potential manufacturers a budget and a supply of raw material, asking them to produce a prototype meeting particular specifications for the company to evaluate.
“In short, we were able to expand our supply chain ecosystem for that product roadmap from one supplier to 22 suppliers, and we were able to qualify 10 of those suppliers with real, physical prototypes,” she says.
Her team has relied heavily on members’ sales skills, and a lot of empathy, when it comes to addressing the concerns that often surround the concept of crowdsourcing.
Those have ranged from engineers wondering whether they’ll effectively be facing outside competition for their jobs to managers skittish about revealing too much of the 130-year-old company’s intellectual property and “secret sauce” to outside interests.
“There was a lot of uncertainty regarding the quality of the outcomes,” she says, as well as how the company could collaborate “with folks around the world you might never meet, in a way that doesn’t disclose any proprietary information.”
The group also deliberately worked early on with a variety of units within GE that Finkhousen anticipated would be particularly open to innovative techniques, picking teams from across the company to help build a diverse portfolio of successful experiments, she says. Before projects even began, the team worked carefully to make sure everyone was on the same page about goals, from commercialization of the end product to confidentiality.
Finkhousen talks about building an “innovation stack” within the company; here’s her slide on what that means:
Creating a Collection of Case Studies
“The results of those early projects became valuable case studies that enabled others to envision how the tools could help them be more successful,” Finkhousen explains. “Knowing these early proof points would help other business and functional leaders understand the value, we curated project types that would resonate with diverse business and functional leaders.”
But as the first successful projects demonstrated the power of expert crowdsourcing, she’s been able to shift gears a bit from internal promotion to handling requests from corporate divisions looking to run their own open innovation projects. She also works closely with the company’s lawyers to make sure that outside participants and GE employees are always clear on the rules of engagement.
“We’re leveraging the same techniques, but our partnership with the business leaders is evolving,” she says. “As our portfolio of successful case studies and business relationships grow, we’re seeing some of our thought leaders evolve from sponsoring individual open innovation projects, to investing in a more sustained and comprehensive engagement of open and collaborative innovation methodology.”
The next step is to help GE’s managers and engineers think of crowdsourcing as a standard item in their toolbox, not just the subject of one-off experiments, says Finkhousen.
“I see the expert community, the solver community, becoming more and more an extension of the GE team,” she says. “I think the GE business leaders will increasingly view the communities as flexible resources that can help us achieve better outcomes [and] that can help us derive better outcomes with our customers.”
Here’s video of a half-hour talk Finkhousen gave in June, at Brightidea’s “Birds of a Feather” gathering in Austin: