Every built system has an architecture; whether we’re talking about mobile payments, power grids, commercial aircraft, or electric vehicles. Thousands, sometimes tens of thousands, of decisions are made to design a product. The field of system architecture works from the belief that there is a foundational subset of decisions that are more impactful than others.
As the director of the System Architecture Group at MIT, Bruce Cameron devotes much of his time to studying and figuring out how to best prioritize the early-stage technical decisions that will determine system performance. Among other things, he and his group have helped architect systems that span from oil exploration platforms for ice-bound drilling to lunar surface exploration vehicles.
And the methods they’ve developed around system architecture are advancing the state of the art. Typical approaches to architectural decisions use prior designs as a starting point for understanding potential designs. But the MIT Systems Architecture Group capitalizes on computational power to design models to explore a host of potential architectures. “Our methods allow us to get a lay of the land before we pick a search direction,” Cameron says. “This gives us greater confidence that the program is going to run on time, finish on time, and run on budget. Frankly, we’re trying to drag product development into the 21st century.”
With Cameron leading, the MIT Commonality Study examined 32 firms over eight years to explore the return on investment on commonality and what tactics firms use to manage the underlying platforms. But one of the big takeaways of the study was that most firms see significant erosion of their commonality targets and weaker ROI when it comes to their platforms.
For example, the U.S. military’s Joint Strike Fighter program worked from the idea that it would be more efficient to build three aircraft on a platform, rather than building three separate aircraft from the ground up. The goal was to share 80-90 percent of the same parts. And the initial cost savings estimate suggested it would be like buying three aircraft for the price of 1.7 aircraft. Twenty-five years later, only 20-30 percent of the parts are shared between the three aircraft, and the process results in the equivalent of buying three aircraft for the price of four.
Cameron and his team found similar behavior and results across industries — from automotive programs to commercial satellites and printing presses. But why does it happen? “From my perspective, there is an inherent tension that exists in a lot of product development,” says Cameron. “Companies want to please their customers by building custom products. But developing one product and leveraging it for a better return on investment is a more efficient use of resources. Platforms embody this tension.”
Through the study, Cameron recognized a need and desire on the part of practitioners to better understand and manage big platforms. As a result, he developed two courses on the subject, one of which is offered through MIT Sloan School of Management, the other through the School of Engineering. The classes have proven immensely popular over the years with more than 1,000 participants, providing Cameron the opportunity to work with practitioners from diverse industries looking to tackle issues around product platforming.
Cameron seems to delight in deconstructing popular paradigms to uncover the truth. “I love ‘the emperor has no clothes’ approach to research,” he says. “It’s not everybody’s cup of tea, it’s a little bit antagonistic, but I think it’s fun, and it’s been a theme of mine for many years now.” To illustrate the point, he mentions the Innovation by Behaving research program, which he developed with one of his graduate students, Owen Ou, now an engineering manager at TotalEnergies, the French oil and gas supermajor.
As with many research endeavors, it began with an observation: “I had lost count of the number of times that I’d heard people yearning to be ‘intrapreneurs’ or wanting to create some version of startup culture within a division of their company,” says Cameron. Having been involved in various startups, he couldn’t help but question the impulse. “Startups are in some sense dysfunctional,” he says, “there are so many reasons that you wouldn't necessarily want to mirror startup culture.”
One of the project outcomes was a list of logic, behaviors, and actions that might benefit a startup but would be detrimental to a larger firm. For example, endless pivoting may not benefit larger firms whose brand equity may be impacted, and encouraging mission-driven behavior that becomes life-defining for the individual can hinder a larger firm’s decision to exit certain markets when the market conditions do not match growth projections.
These days, Cameron has been particularly interested in research around the rapidly evolving satellite communications market: the players, its past, and future direction. Much of the work, he says, began with an MIT ILP-brokered introduction to a major satellite operator. The firm was looking for a strategy to compete in a market newly flooded with supply thanks to billionaires launching satellite networks. Reflecting on his relationship with the Industrial Liaison Program, he says, “MIT Industrial Liaison Program has a hard job, helping outsiders navigate the Institute. But I've been privileged to work with ILP throughout my career. Frankly, it's been some of the most fun that I've had at MIT. I think MIT can be quite a challenging environment to operate in, and I've been really pleased with the relationships ILP has initiated for me over the years.”