When iRobot filed for bankruptcy on Sunday, the immediate explanations were familiar: cheaper competitors, margin pressure, a failed Amazon acquisition, rising costs. All true — but incomplete.

The more interesting story isn't that iRobot failed to innovate. It's that it innovated inside a system that quietly trained it to stop running — a textbook case of the innovator's dilemma, amplified by patent protection.

The sedative effect

For much of its life, iRobot sat behind an unusually strong patent moat. By around 2017, the company held roughly 400 US patents and more than 900 worldwide — a portfolio ranked in the top tier of IEEE's patent power scorecard for electronics. Its core technologies — navigation, cleaning mechanisms, docking behaviours, mapping — were legally protected and aggressively enforced.

This wasn't passive protection. iRobot went to war. Between 2017 and 2022, the company filed International Trade Commission complaints against Bissell, Hoover, bObsweep, iLife, Black & Decker, and SharkNinja. It won exclusion orders barring competitors' products from the US market. It forced settlements. It made copying expensive and legally risky.

And here's the uncomfortable truth: it worked. For a while.

Patents didn't just keep competitors out — they bought iRobot something far more subtle and dangerous: time. Time to move slowly. Time to premium-price. Time to assume leadership was stable.

Protection, in effect, muted competitive pressure. And sustained pressure is what keeps organisations sharp.

The cliff edge

Patents don't decay gradually. They expire. And when protection collapses, it collapses completely.

As iRobot's key patents began to run out, the competitive dynamic changed abruptly. Rivals didn't cautiously test the waters; they flooded the market. Features that once justified Roomba's premium pricing — laser navigation, room mapping, auto-evacuation, mopping — appeared everywhere, bundled together, and sold for less. Ecovacs. Roborock. Dreame. Companies that had been blocked or deterred were suddenly free to compete on features iRobot had spent years defending.

The failure wasn't that patents expired. That was inevitable.
The failure was that iRobot hadn't built the muscle memory to compete without them.

By the time the company responded, the ground had already shifted. Its newest products finally added features competitors had standardised years earlier. But these weren't category-defining moves — they were survival moves. The most telling detail is that recent Roombas barely look like Roombas at all. Outsourced designs, stripped-down engineering, and margin-first decisions replaced the distinctive solutions that once set the brand apart.

This wasn't innovation. It was adaptation under duress.

The Amazon illusion

There's another layer to this story. In August 2022, Amazon agreed to acquire iRobot for $1.7 billion. For a company under competitive siege, this looked like salvation — absorption into Amazon's ecosystem, access to vast resources, a future beyond the patent cliff.

Except European regulators killed the deal in January 2024, concerned that Amazon could foreclose rivals by restricting marketplace access. Amazon paid a $94 million breakup fee and walked away. Colin Angle, CEO since iRobot's founding in 1990, resigned soon after.

The collapse matters because it helps explain why iRobot's patent-centric strategy persisted for so long. When you believe a $1.7 billion exit is coming, you optimise for acquisition, not competition. You preserve margins rather than slash prices. You defend territory rather than expand capability. That belief quietly reshapes incentives inside the organisation.

The Amazon lifeline made the patent strategy seem rational — right up until it wasn't.

A different operating system

To understand why iRobot struggled once protection disappeared, it helps to look at an ecosystem built without it.

In Shenzhen and the Pearl River Delta, innovation runs on a manufacturing philosophy Western companies often misunderstand. Known as shanzhai — literally "mountain fortress," a reference to operating outside imperial control — it describes an ecosystem where copying isn't theft but a starting point. The real competition happens in what comes next.

Silvia Lindtner, a University of Michigan professor who has studied Shenzhen's manufacturing networks extensively, describes the core advantage: companies put small-batch runs of a few thousand devices on the market to see if they fly. If they do, they scale. If not, they iterate. The cycle time is measured in weeks, not years.

The system is built on openness. Manufacturers share reference circuit boards called gongban — public bases that provide standard functions like Bluetooth connectivity, sensors, and basic computing. Anyone can modify them. Anyone can build on them. The assumption isn't that your design will be protected — it's that it will be copied, probably within months. Your only defence is to move faster than your imitators.

This creates a fundamentally different innovation culture. In traditional Chinese craft training, copying and repetition aren't failures of originality — they're respected steps toward mastery. Frequent iteration and incremental improvement are seen as necessities on the path to excellence. Only later is originality encouraged.

The result is companies that treat competitive advantage as temporary by design. Product cycles are short. Hardware and software evolve together. Execution matters more than elegance. Leadership is something you earn continuously, not something you defend legally.

What killed Nokia

Here's the part that should worry any company relying heavily on patent protection: the shanzhai ecosystem didn't just compete with Western incumbents. It hollowed them out.

When Nokia and Motorola dominated mobile phones in the mid-2000s, they sold feature phones for $800 to $1,000. They had patents. They had brand recognition. They had distribution. What they didn't have was a response to manufacturers selling functional phones for under $100.

The conventional narrative says Apple killed Nokia. But the smartphone transition was only part of the story. Shanzhai manufacturers had already eroded Nokia's volume base in Africa, Latin America, Southeast Asia, and India before the iPhone appeared. By 2010, shanzhai phones accounted for roughly 20% of the global 2G handset market.

Nokia's patents didn't save it. They delayed recognition that the game had changed.

The final irony

iRobot's bankruptcy filing contains a detail so symbolically perfect it barely needs commentary.

The company is being acquired by Shenzhen-based PICEA Robotics — its primary contract manufacturer and secured lender. Under the restructuring plan, PICEA will take 100% ownership. Existing shareholders will be wiped out. The debt will be cancelled.

The company that made iRobot's products under licence now owns iRobot outright.

This is what happens when you outsource manufacturing to stay asset-light while relying on patents to stay competitive. Eventually, your manufacturer learns everything you know, builds the relationships you need, and accumulates the leverage you lack. When the music stops, they're holding the chairs.

Twenty years is a long time in technology

There's a policy dimension to this story worth noting. US utility patents last 20 years from filing — a term designed for a world of heavy manufacturing, where inventors needed time to build factories, establish supply chains, and recoup large capital investments before competitors could legally copy their work.

That logic makes sense for pharmaceuticals or industrial machinery. It makes far less sense for fast-moving consumer technology.

iRobot's foundational patents were filed around 2002–2003, when the original Roomba launched. Core protection began expiring around 2022–2023 — precisely when competitive pressure peaked. Twenty years of insulation. But robotics, sensors, and software changed beyond recognition during that period. The patents outlasted their technical relevance while blocking competitive fitness the entire time.

Here's the counterintuitive possibility: shorter protection might have been a gift.

If iRobot's foundational patents had lasted ten years instead of twenty, the competitive cliff would have arrived around 2012–2013 — when the company was still strong, founder-led, and capable of adapting. Instead, long protection allowed it to coast into a reckoning it was no longer equipped to survive.

In fast-moving markets, first-mover advantage often provides sufficient protection on its own: brand recognition, ecosystem lock-in, accumulated expertise, iteration speed. By the time competitors copy you, you should already be two generations ahead. If you're not, patents are just delaying the inevitable.

Two theories of competitive advantage

The iRobot story illustrates a strategic choice that extends far beyond robot vacuums.

The patent model says: build walls, litigate, slow the game down. Raise the cost of competition. Use the time you buy to extract premium margins and optimise known products.

The shanzhai model says: share foundations, assume copying, compete on speed. Treat every advantage as temporary and every product as a stepping stone.

iRobot was playing the first game in a world increasingly governed by the second.

This pattern repeats. Kodak and film. Nokia and platforms. BlackBerry and keyboards. Blockbuster and distribution. In each case, protection — legal, structural, or market-based — delayed adaptation rather than enabling it. The walls felt like strength. They were actually weight.

The takeaway

Patents are meant to protect innovation. But when they become a substitute for urgency, they can do the opposite. They encourage optimisation over reinvention. They reward being right yesterday. They slow organisations just enough that, when protection disappears, the race is already lost.

The real warning in iRobot's collapse isn't about competition, China, or patents themselves. It's about mistaking insulation for strength.

Companies that assume their advantages will last tend to build slower muscles. Companies that assume they won't tend to build faster ones. When the environment changes — and it always does — only one of those survives.

Innovation doesn't just need freedom.
It needs pressure.

And the most dangerous moment for any company isn't when competitors appear — it's when protection makes them feel like they don't matter.