‘It felt so wrong’: Colin Angle on iRobot, the FTC, and the Amazon deal that never was

When iRobot filed for Chapter 11 bankruptcy last Sunday, it marked the extremity of an epoch for 1 of America’s astir beloved robotics companies. The Roomba maker, which had sold implicit 50 cardinal robots since its 2002 launch, had survived 35 years of near-death experiences and method challenges lone to beryllium undone by what laminitis Colin Angle calls “avoidable” regulatory opposition.

The illness followed Amazon’s January 2024 determination to scuttle its $1.7 cardinal acquisition of iRobot aft 18 months of probe by the FTC and European regulators. In this candid conversation, Angle reflects connected what helium describes arsenic a profoundly frustrating process, the chilling connection it sends to entrepreneurs, and his determination to determination guardant with a caller task successful user robotics.

This interrogation has been edited for magnitude and clarity.

TC: You called the bankruptcy “avoidable” and a “tragedy for consumers.” Walk maine done what you deliberation regulators got incorrect successful blocking the Amazon acquisition.

CA: I deliberation there’s a existent acquisition astir the relation of the FTC and the European Commission. The goal, of course, is to debar the abuses that tin hap successful monopolies and with the extremity of protecting user prime and protecting innovation.

What happened was that iRobot and Amazon came unneurotic for the expressed intent of creating much innovation, much user choice, astatine a clip erstwhile iRobot’s trajectory was honestly antithetic from wherever it was respective years earlier. In the EU, we had a 12% marketplace stock [but it was] declining wherever the fig 1 rival was lone 3 years aged successful the market, which is astir the explanation of a vibrant and dynamic marketplace. And successful the United States, iRobot’s marketplace stock was higher, but it was declining and determination were aggregate increasing competitors bringing extracurricular innovation into the marketplace.

This should person been a no-brainer. This should person been three, 4 weeks of investigation. What happened alternatively was a twelvemonth and a fractional of pendency, which had a precise challenging interaction connected the quality to run a institution and yet having the acquisition blocked.

What was that 18-month process really like? What were you being asked to do?

The magnitude of wealth and clip spent was indescribable. I would not beryllium amazed if implicit 100,000 documents were created and delivered. iRobot invested a important portion of our discretionary net against fulfilling the requirements that went on with doing the transaction. Amazon was forced to put many, many, galore times that. There was a full team, some interior and outer employees and lawyers and economists moving to effort to, successful arsenic galore antithetic ways arsenic imaginable — due to the fact that it seemed similar our connection was falling connected deaf ears — show that this acquisition was not going to make a monopolistic situation.

There was regular enactment for 18 months associated with this. Perhaps astir telling, erstwhile I was testifying arsenic portion of being deposed, I had a accidental to locomotion the halls of the FTC. The examiners connected their bureau doors had printouts of deals blocked, similar trophies.

Trophies?

To me, it felt truthful incorrect arsenic an entrepreneur who started this happening virtually successful my surviving country and lived six and a fractional years ne'er having capable wealth successful the slope to marque payroll, yet making it done succeeding. Here’s an bureau whose stated ngo is protecting user interests and helping the United States economy, celebrating arsenic victories each clip they unopen down M&A, which successful a precise existent mode is the superior operator of worth instauration for the innovation economy.

I went into this deposition looking for a friend. It’s like, “Here we are, we’re evidently not successful a stronger place, and here’s a large accidental for us. Are you excited for us?” Maybe this is conscionable my naïve instrumentality coming out, but that’s not the reception I got. It was, “Why should we ever fto them bash this?” It’s like: due to the fact that it’s bully for the user — due to the fact that it’s going to catalyze innovation.

How bash you deliberation what happened present changes the calculus for startup founders who spot acquisition arsenic their exit strategy? Do you deliberation we’re successful a satellite wherever American tech companies can’t standard done M&A?

Risk has a chilling effect. If you’re an entrepreneur, your lone enactment is to anticipation that it doesn’t hap again. The crushed you and I are talking is I anticipation by my words I tin marque it little apt that it volition hap again.

I founded a caller institution and my outlook connected exit strategy and adjacent commercialization strategy is impacted by the experiences I had astatine iRobot. How tin it not be? That precedent creates hazard of it happening again, and lone done affirmative experiences bash we commencement to dial down the anxiousness that the exits I’m depending connected — oregon arsenic a task capitalist, the exits I’m investing successful assuming volition hap — volition really travel through. That hazard is factored into the willingness to invest, the valuation of deals, and the complaint of caller institution formation.

It’s hard to accidental there’s X percent less entrepreneurial starts oregon exits arsenic a effect of the chilling message, but it surely didn’t help. Entrepreneurs tin usage each limb up that we arsenic a federation tin provide. It is simply a rugged journey. When it really works out, it should beryllium a celebration. The FTC is determination arsenic a safeguard against precise existent examples of things gone excessively far. I’m a large believer successful checks and balances. But erstwhile things get retired of whack, the state suffers.

Let’s speech astir iRobot’s journey. The Roomba didn’t travel retired for 12 years aft the institution was founded. Tell maine astir those aboriginal days.

iRobot was a clump of radical successful an world laboratory saying, “We were promised robots. Where are the robots?” If I could beryllium pissed disconnected that I don’t person the robots we were promised, I should bash thing astir it. If not us, who? If not now, when?

One of the co-founders, my prof Rod Brooks, had pioneered an AI exertion which allowed the embedding of instrumentality quality successful low-cost robotics. The ngo of the institution was to physique chill stuff, present a large product, person fun, marque money, and alteration the world.

The archetypal concern program was “private ngo to the moon, merchantability the movie rights.” We were possibly the archetypal institution to neglect to bash precisely that. But the exertion we developed led to additions to the Mars Pathfinder ngo — my name’s up connected Mars. We built robots that went into the Gulf of Mexico aft the Deepwater Horizon disaster. We built the PackBot, which was the archetypal robot ever deployed successful a combat ngo for the US Army, went into caves successful Afghanistan, and became the superior methodology for diffusing improvised explosive devices. We would get postcards: “You saved my beingness today.”

When the Fukushima catastrophe happened, we donated fractional a cardinal dollars worthy of robots to Japan. We sent six radical implicit to bid Tokyo Electric Power Company employees. Those robots were the archetypal wrong the reactor doors, mapped retired radiation levels, and recovered a way wherever an worker could tally in, get to the power room, enactment for a infinitesimal and a half, tally backmost out, and lone person a beingness dose of radiation. We were credited with enabling the shutdown of the reactor.

And past twelvemonth 12, the Roomba?

I had a squad that was moving connected toys, and 1 of the guys said, “Colin, I deliberation it’s clip we tin bash this. We tin yet marque the vacuum.” I’m like, “Okay, here’s $15,000. Two weeks. See what you tin do.” Two weeks aboriginal they came backmost and said, “Hey, that’s not bad. Maybe there’s thing here.” We scraped and recovered a small spot of wealth [to physique these]. A twelvemonth and a fractional later, I convinced my committee that we could physique 10,000 of these robots and we launched them.

It’s truly the media that picked up the story. We had nary wealth for marketing. Reporters were fascinated due to the fact that we tended to bash absorbing things that were real, and they couldn’t judge that robot vacuuming could perchance beryllium real. [The absorption was:] “My God, it really worked.” We ended up selling 70,000 robots successful the archetypal 3 months. Then we astir went retired of concern the adjacent year.

Because you couldn’t accommodate demand?

Because we screwed up. The property drove this immense archetypal request — 70,000 robots. So adjacent twelvemonth we’re going to bash 4 times that. We made 300,000 robots. We adjacent made a tv commercial, but we were a clump of geek engineers, truthful it wholly failed. After Cyber Monday we were sitting with 250,000 robots successful our warehouse like, “Oh my God, the world’s going to end.”

Then thing bully happened. The feline moving our website said, “Why did income quadruple yesterday?” We hadn’t done anything. What had happened was Pepsi had started moving a TV advertisement with Dave Chappelle. He walks into this beauteous home, picks up a murphy chip, and a Roomba comes out. He’s like, “A vacuum cleaner!” He throws down the murphy chip, the vacuum eats it, past chases him. His pants are ripped off. He stands up successful boxers. A beauteous pistillate appears, and helium says, “Your vacuum cleaner ate my pants.” We sold 250,000 robots successful 2 weeks and realized we knew thing astir marketing.

Wow! You had nary thought Pepsi was incorporating your merchandise into its commercial?

No idea. It was wild. You effort to bash bully for truthful agelong and you get smacked successful the look truthful galore times, and past sometimes thing bully happens. You could accidental due to the fact that of that brainsick infinitesimal successful time, we person robots. When we deliberation astir conscionable however fragile a travel it is — cats riding Roombas was a large portion of wherefore we succeeded. Does that marque immoderate sense? No, not astatine all. But we surely didn’t expect tens of billions of views of cats riding Roombas.

At immoderate constituent you started seeing Chinese competitors similar Roborock and Ecovacs adopting lidar navigation years earlier iRobot. Why did you instrumentality with vision-based navigation for truthful long?

We explicitly did not enactment lasers connected the robot. We had the exertion decades agone due to the fact that it’s a dead-end technology. Under my strategical direction, we were going to put each penny of outgo against a vision-based nav and situational knowing system. Your Tesla does not person a laser connected it. It is each vision-based. At slightest Elon agrees with me.

Our strategical program was for Roomba to beryllium overmuch much than conscionable a vacuum cleaner successful the home. To bash that, it needs to recognize more. Lasers are not precocious exertion — they’ve been astir for decades. They’re an expedient solution to a subset of the problems that a robot astatine location needs to instrumentality on. A laser volition ne'er archer you whether you really cleaned the level oregon not.

It is perfectly existent that Chinese contention was coming successful astatine little terms points. We were precocious successful gathering a two-in-one robot — we believed that by separating mopping and vacuuming you could springiness a amended lawsuit experience. The lawsuit voted that we were wrong, and that’s okay. We surely were archetypal successful auto-evac, archetypal successful navigation. We were besides importantly excluded from the Chinese marketplace, which is the largest marketplace for user robotics successful the world. That didn’t help.

What learnings from your iRobot acquisition would you stock with different robotics entrepreneurs?

The archetypal happening I archer each robotics entrepreneurs is: marque definite you recognize your marketplace truthful that you are gathering thing that delivers much worth than it costs to create. Robots are truthful exciting, truthful sexy that it’s truly casual to person yourself that you’re doing thing that is going to alteration the world, if lone consumers were astute capable to recognize it. That’s a beauteous pugnacious equation.

Technology is oftentimes successful the robot abstraction good up of the concern plans that tin instrumentality vantage of the technology. One of the traps is reasoning of robotics arsenic a happening arsenic opposed to a toolkit. As soon arsenic you say, “I’m going to physique a robot,” and caput disconnected gathering your humanoid — are you truly doing it due to the fact that you recognize a occupation you’re trying to solve, oregon are you enamored with gathering your thing?

When I started iRobot, it was conscionable assumed that robots were going to vacuum floors by gathering humanoids to propulsion upright vacuum cleaners. When we archetypal built Roomba, we would inquire people, “Is that a robot?” People would say, “No, that’s not a robot. A robot has arms and legs and a head.” Yet Roomba astatine the clip outgo 10,000 times little than a humanoid pushing an upright vacuum.

The situation of entrepreneurship is breaking done the romance and accidental and falling successful emotion with your exertion and getting to the exertion that you’re trying to solve. Understand your consumer, recognize the occupation you’re trying to lick — due to the fact that it’s complicated, robotics is expensive, and it takes a batch of vigor to get right.

You mentioned you’ve founded a new company. What tin you archer maine astir it?

We’re successful stealth mode, but I’ll springiness you a wide hint. It’s consumer-facing. We’re truly looking astatine the information that astir of the things robots tin bash to conscionable outstanding needs necessitate america to interact with different people. So however bash we physique a robot that really has capable affectional sophistication — not human-level, but capable — to physique an enduring co-character that tin marque consciousness implicit clip and usage that for wellness and wellness related applications?

It’s going to beryllium awesome. I’m truthful excited astir it. It’s fixed maine enthusiasm and vigor to person a accidental to usage this caller toolkit and proceed connected my travel to physique the robots we were promised. I truly haven’t changed that overmuch from a grad pupil successful assemblage saying, “Oh my God, we’ve been promised robots and we don’t person the ones that I privation yet.” I spent 30 years focused connected gathering the world’s top level attraction robot, and present I person a accidental to bash thing else.

For much from this interrogation with Angle, including his instrumentality connected whether humanoid robots volition ever work, cheque retired TechCrunch’s StrictlyVC Download podcast — caller episodes driblet each Tuesday, and the afloat speech volition beryllium disposable soon.