Moving from a push to pull system can improve product timelines 

Don’t curse the next time your check engine light comes on — credit a mind like Oljeta Qirko for streamlining the process of getting smart sensors from the drawing board to your dashboard.

As part of her Organizations Lab work, Qirko introduced a workflow pilot into an innovation process in her office at smart sensor company, cutting the time from ideation to minimum viable product by 60 percent. In Organizations Lab, MIT Sloan Executive MBA students undertake a change initiative in their companies, diagnosing a problem and leading improvement. Faculty and peers voted Qirko’s work as the Organization Lab winner. She and another finalist presented their work to peers, faculty, and staff at a recent EMBA ’18 and ’19 joint dinner event.

“Our innovation process is designed to be a fast learning and iteration loop, through early customer engagement,” Qirko, EMBA ’18, said.

In this case it means sensor developers are getting that much closer to matching the speed of safety and security features with the rapidly changing automotive industry. But Qirko’s pilot can apply to a range of organizations looking to move their development workflow from a push to pull system — a method used by well-known corporations like Toyota.“If I already know what tasks are mine, I don’t have to wait for someone else to get started,” Qirko said. “That creates a pull system.”

The automotive industry is going through major changes thanks to electronic cars, autonomous vehicles, and ride sharing, Qirko said. The sensors that her company creates are used in the aerospace and automobile industry — harsh environments that use electronically complex systems.

“We need to be ahead of the curve with our product technology and quality, which requires yet another level of rapid innovation and design agility,” Qirko said.

Qirko said the smart sensor company she works for started a new innovation process about a year-and-a-half ago, with the goal of a more agile approach to projects. After a year observing the new innovation workflow, Qirko came to the conclusion that there was room for improvement.

She found instances of misalignment on growth opportunities pursued, delays caused by communication disconnects between internal teams, problems with under-scoped projects, and teams stretched across too many projects.

“We were entrepreneurial and worked hard to get working prototypes in the hands of our customers, as soon as we could,” Qirko said. “However, innovation is messy and iterative and our existing product development process was not designed to handle it efficiently.”

Other problems arose in how projects were handled in a series, rather than in parallel tracks, Qirko said. A prototype of a product would be drawn up and manufactured, tested, and then sent to the customer, rather than engaging users earlier in the process to ensure the right product was being made to serve a need.

“In several examples, we overdesigned,” Qirko said. “We were so focused on developing a product, instead of a learning tool. At times, we produced a sensor prototype that was way more than we needed for a rough prototype. We could have been much faster to market had we followed an iterative approach.”

But these problems were to be expected in a grassroots effort, Qirko said. To address them, Qirko created a pilot program that cut the 11 steps of the current workflow down to five.

The new workflow started with an all hands meeting at the beginning of the project, rather than a back and forth throughout the different innovation stages. Decision makers from sales, marketing, and engineering gathered right from the get go to lay out the team and scope of work.

The pilot program dedicated team members to one project. Using visual management (think walls of Post-it notes), team members could see the flow of work and their responsibilities. The pilot also encouraged the use of single-sheet reporting, so that everything from the problem statement to planning to customer engagement were all in one place.
In practice, the original innovation process time dropped from 20 weeks to eight weeks.

Qirko said her pilot project received support from company leadership, and her office has already started to adopt the one-sheet and visual management practices.

While she does wonder how the productivity gains will be maintained as more project opportunities come to the team, Qirko hopes the lessons learned from these pilots become part of the company culture.

“Ideally instead of a pilot,” she said, “it becomes the natural way of doing our work.”