Commercial Floor Cleaning Robots from Brain Corp

July 30. 2017. 6 mins read
Table of contents

Those of us who are old enough to remember the dot-com days know what it feels like at the peak of technology hype. Maybe we’ll never see a state of irrational exuberance like that again, but at least we had a chance to see what real hype looks like. The Gartner Hype Cycle refers to this as the “peak of inflated expectations”, which is usually followed by the “trough of disillusionment” where we realize that this great technology hasn’t delivered on its promises of greatness. Everyone panics, lots of selling takes place, and people forget about it for a while before it comes back around and begins to do great things.

Let’s take artificial intelligence (AI) as an example. We wrote before about how artificial intelligence is not like the dot-com era because “this time it’s different“. Startups with bold claims are having to show their work in order to demonstrate that they are actually using AI to achieve a competitive advantage. With +3,000 AI startups in the game and growing, it seems like that notion is being thrown to the wayside. You can tell that AI technology is reaching critical mass because people are starting to issue drastically differing opinions about where the whole thing is heading. According to some, AI has already began developing a conscience and is starting to exhibit such charm that it’s being asked out on dates:

Then, not more than 24 hours passes before we see this riveting piece of news from the New York Times:

With all the mixed messages we’re getting, how can we tell if AI is real or not? The answer lies in the numbers. You have to address each case individually and look for numbers that demonstrate AI is doing something remarkable. Because AI is literally built on data, if people can’t show you quantified results then they’re full of isht.

Take AI startup Afiniti as an example. They’re so confident in their numbers that they don’t even charge for their product. They simply want a cut of what they save their clients. Another startup called Boomtrain (which was just acquired) can perform email marketing optimizations that are beyond what any human can do with A/B testing. In both of these cases, the value on offer is demonstrated in the numbers.

In other cases like autonomous driving, the potential doesn’t need much explaining. That’s because the labor saved by replacing 3.5 million human truck drivers is enormous, not to mention the decreases in liability and savings on insurance due to the safety of autonomous driving. Massive amounts of money are being thrown at the idea in hopes of backing the right dog in the race and producing “the next Microsoft”. These startups pursue a land grabbing strategy that looks to scale faster than anyone else in order to capture the majority of market share. When we evaluate startups like these, we need to look at who is backing them. One startup called Brain Corp is looking at using AI powered autonomy to wreak some havoc in spaces that aren’t subject to regulatory risk.

About Brain Corp

Click for company websiteFounded in 2009, San Diego startup Brain Corp has taken in $125 million in funding so far, which up until a few weeks ago had entirely come from a $78 billion chipmaker called Qualcomm (NASDAQ:QCOM). Then, the biggest investor in technology today, Softbank, decided to invest $114 million into Brain Corp this month alongside QCOM. Given Softbank’s successful track record of investing in technology, and Qualcomm’s expertise in hardware, it’s likely that the Brain Corp’s technology has been thoroughly vetted.

Update 04/24/2020: Brain Corp has raised $36 million in Series D funding for further expansion into retail, healthcare, airports, and schools. This brings the company’s total funding to $161 million to date. 

So what does Brain Corp do?

Brain Corp is building something called EMMA which stands for Enabling Mobile Machine Automation. EMMA is an AI navigation system that’s “capable of automating a range of commercial-grade equipment – everything from industrial floor care machines, to medical equipment and forklifts“. In particular, Brain Corp is focused on commercial floor cleaning robot like this one:

Floor Cleaning Robots
Source: Brain Corp (Our emphasis in yellow)

The above commercial floor cleaning robot is equipped with sensors, 4G connectivity, cameras, and the machine learning algorithms that learn from a human who cleans the store the first few times. After that, AI “frees them up to focus on more value added activities” and then makes its way around a store while avoiding any obstacles that pop up like boxes left on the floor or even humans. Considering that labor accounts for 80% of the costs of commercial floor cleaning, you can imagine how much this is making low-margin retailers salivate. It’s also being perceived as a real threat by labor unions, something that Brain Corp is proactively getting in front of by sidling up to the International Sanitary Supply Association (ISSA).

As of 2016, there were 2,161,740 janitors working in the USA. If just 10% of them are “freed up to focus on more value added tasks” then based on the mean salary of janitors in the U.S., that would equate to $5.2 billion in savings. If you’re a company that sells floor cleaning equipment to commercial property managers, you suddenly have a very compelling value proposition for your clients. Brain Corp wants to enable those companies that sell traditional floor cleaning machines with the ability to enable their machines with autonomy, and they’ve already done this:

Floor Cleaning Robots

The way that Brain Corp makes money isn’t just the licensing of their tech to be used in these expensive machines, like the ICE RS26 which retails for $18,000. The real opportunity is in the recurring revenues they expect to capture from the end user. Here’s how it works:

  • Manufacturers integrate Brain Modules into their branded equipment.
  • Customers purchase Brain-enabled equipment from manufacturers.
  • Customers activate Brain Corp’s monthly service to automate their equipment.

This means that Brain Corp can now rely on recurring revenue from the end customer who is eager to use a service that helps them save on labor costs. They can then use what they’re learning from the floor cleaning industry and then start to attack other industry verticals. In fact, the reason they raised this money was “to further develop A.I. technology and create brains for multiple types of commercial and consumer robots“. Even Masayoshi Son, the Chairman & CEO of SoftBank, sees the big opportunity here when stating that “Brain is developing truly ground-breaking technology that transforms manually operated machines into autonomous robots“. That core technology is called BrainOS:

The reason BrainOS is different is that you don’t need to deploy beacons around your store or use GPS which can be spotty. The system first learns from a human and then constantly learns over time. For example, it may learn that your stocking crew (who are probably next on the list to be “freed up”) always start with stocking the soda aisle every night at 11 PM, and that on Friday/Saturday nights the beer section is busy with humans until 2 AM.

Brain Corp has dropped hints about moving into medical equipment and forklifts. Regular readers will recall an article we wrote before titled “Seegrid – Are Robot Forklifts Stealing Jobs?“, where we looked at a startup that’s taken in $53 million and has a client list with big names like 3M, Volvo, Walgreens, Whirlpool, and even the U.S. Postal Service. Like Brain Corp, Seegrid robot forklifts operate by “sight” and don’t need to be guided by lasers, wires, tape, or beacons.


There are probably many other startups out there as well given the size of these opportunities. One would assume that being the expert investors that they are, Softbank vetted all the technologies on the market today and figured Brain Corp and their BrainOs was the best bet for these particular applications. It seems that contrary to the clickbait headline from New York Times, artificial intelligence is far from stuck.


Leave a Reply

Your email address will not be published.