The Factory OS
What American Manufacturing Can Learn From Uber, Oscar and Nest
Say what you will about start-up valuations, we’re living through a golden age of product design.
Not only are our apps getting faster, cheaper and more beautiful, but they’re branching off into the physical world. Our favorite new applications, from Uber to Nest, are applying the best lessons of the consumer web to real-world products and experiences.
While they might share a home screen with Snapchat or Twitter, these apps are closer to operating systems: controlling a range of applications and hardware through a common interface.
Instead of disk drives and CPUs, Uber’s TransporationOS runs on drivers, black cars and smartphones; managing a suite of applications that let passengers book rides, track their driver’s location and pay from their smartphones. Together they form a unified experience that starts and ends in Uber’s smartphone app, but could span an entire city. In a deeply human and unpredictable world of road rage and rush hours traffic, Uber manages to close the loop.
Uber’s not alone. In fact, it’s just one of a growing number of real-world operating systems closing the loop on everything from healthcare (Oscar), to self-storage (Makespace), to late-night alcohol delivery (Drizly) that are making the world a little healthier and less cluttered with every install.
However, there’s one major industry that has been untouched by app developers, despite its size and importance to the economy — manufacturing.
While manufacturing makes up more than a fifth of the US economy, its “operating system” hasn't changed since the Reagan administration. At a time when both the technology and the market for manufacturing are exploding, isn't it time we rethought how the pieces fit together?
When we started Matter, we asked ourselves what a factory operating system might look like. What if the end-to-end process — from tooling, to sourcing to shipping — could be controlled through a common interface?
We thought about what lessons to take from the world of product design, that could help us close the loop on manufacturing. They would be the killer apps and features of a FactoryOS:
One: Real-Time Information (Uber)
The real magic behind Uber’s operating system is real-time information. Within seconds of opening the app you can: locate nearby cars, get a quote, book a ride and get a down-to-the-minute ETA of when your driver will arrive. Whether you want to upgrade cars or it starts to rain, those numbers instantly change. By bringing real-time information to the center of the user experience, Uber has taken the frustration and guess work out of getting around.
While waiting for a cab is annoying, waiting for a production run can be deadly. For designers and small businesses who manufacture, information — about pricing, production and turn-around time — is categorically opaque. At its best, working with a factory is mediocre and at its worst, can completely derail a businesses, making it impossible to meet demand.
The longer it takes to get that information and the less complete it is, the harder it becomes to grow a business that relies on manufacturing. If designers could see into their factories with the same clarity Uber passengers can view their rides, production could be an asset rather than a liability.
Two: Smart Filtering (Oscar)
While Uber is a great model for simple purchasing decisions, more complex products require an extra layer of architecture. Next to manufacturing, health insurance is a poster child for complexity, but the start-up Oscar has done a great job of making the process of buying and using insurance simpler.
Like many insurers, a visit to Oscars’ website will prompt you to get a quote, but before seeing a single plan or detail, Oscar collects key variants — age, dependents, income, location — which it uses to personalize your experience, delivering the most popular products and personalized pricing based on your profile.
Once you’re an Oscar customer you can: book an appointment, explore your plan, or get a consult. Beneath that, there’s a web of connected information — from common ailments and the effects of different drugs, to the nearby doctors who might prescribe them, all of which is accessible through a large search bar at the top of every page. By intelligently filtering information, Oscar makes complex decision-making simple.
To the non-expert, manufacturing is every bit as complex as health insurance. Similarly, the kind of information you need to make decisions widely varies depending on what you’re making and what services you require. Organizing options around specific projects and customers could save designers a lot of heartache and make manufacturing accessible to a new generation of small businesses.
Three: Leverage What’s Local (AirBnB)
While apps like Uber try to replicate the same service wherever the user goes, others, like AirBnb, make regional differences a key part of their offering. Traveling to Europe over Christmas? Try a Parisian apartment overlooking the Seine. If it’s South America in summer, you can find a bungalow at the base of the Andes.
Like hospitality, manufacturing has different local advantages and characteristics. There are regions famous for their goldsmiths and stone cutters, while others are known for their supply of cheap and reliable labor. In a manufacturing operating system, the location of your factories should be a function of what you’re making — and should change as quickly as you can tweak your design or change a batch size.
Four: Continuous Learning (Nest)
One of the key advantages software-based services have over their real-world counterparts is their ability to quickly aggregate and learn from data. Machine learning improves outcomes through algorithms rather than relying solely on human analysis.
Learning is a central feature of products like Nest’s thermostat, which gets to know your climate preferences and optimizes them to conserve energy. Nests’ climate control operating system does this automatically without users having to program it or even understand their own behavior.
A factory that learns like a Nest thermostat could have powerful implications for manufacturing — improving cost, quality, and speed with each production run. It could learn about different structures to make smarter design recommendations.
By monitoring the nuances of performance across factories it could find the best fit for a given item or run and identify best practices which it could ‘install’ across the network as settings on the factory floors. If it were tied into inventory, it could be even smarter, producing different sized batches, in different locations, based on the velocity and distribution of sales.
If chunky belt buckles are trending at Sundance, a boutique could overnight a batch from nearby in Brighton. Or, if espadrilles are taking off in the south of France, a retailer could stock its European stores with them — all at the prompt of an intelligent operating system.
