What robotics investors say they’re looking for in startup pitches | Hacker News Reader

What robotics investors say they’re looking for in startup pitches

techcrunch.com

64 pointsjanober31 comments

Open articleView on HN

What are the properties of successful robot startups?

- iRobot - made enough on military contracts to enter the vacuum business.

- Kiva - new, low-cost approach to warehouse operation. Changed fulfillment industry. Bought by Amazon.

Those are about the only recent startups I can think of that made it.

Startups not doing too well:

- Rethink Robotics - Rod Brooks' startup. Heavy on the hype, light on the profits.

- Abundant Robotics - apple-picking machine. VC financing, yes. Production hardware shipping and being used, no.

In industrial robotics, the successes are all old-line industrial companies - Fanuc, Epson, Comau, ABB, and Foxconn.

I'd argue DJI is a robotics company, although there are a lot of definitions you could use (you could also say self-driving cars are robots, in which case Cruise was a financial success).

I think your point remains though. Robots really epitomize hype over profits. It's a tough industry where people ignore margins while trying to build impossibly-difficult tech.

I agree that DJI is a robotics company, and as somebody who has been keeping a close eye on the consumer drone market, it seems they are practically impossible to compete with.

DJI is a great consumer robotics company, but they're built on top of a mass amount of experience gained from the DIY world, and this can't be underestimated for its residual effects .. sure, you can buy a DJI consumer product anywhere for a few grand, but for the price of a single DJI you can build yourself a small fleet. And I think that is where they are going to get their competition from, ultimately .. the RCGroups.com crowd.

I dunno, I am doing exactly this right now. Rather than by a DJI spark for £500 I'm building my own micro drone.

current expenses are at about £100 for the frame, motors, ESC, and flight controller, but I already own a transmitter and receiver that I'm reusing, so add another £100 minimum for that (probably more tbh). Then factor in that so far I've not brought the LiPo batteries or chargers, so add another £50. Then factor in I'm not putting any cameras on it yet, so add another £100 for a fpv camera, 5ghz transmitter and receiver. And the spark does more than that would provide, to match the spark you'd also need to add a second HD camera, and gimbal, so maybe another £100 if you buy low cost parts (I'm not currently seeing any diy parts that can do what the spark does I.e live view and recorded HD video with one camera, hence the two cameras needed). So that's £450 now.

And then consider that the spark comes with a smart phone app, GPS, and a bunch of nifty easy to use features that are way ahead of what clean flight/betaflight open source controller software currently does.

Perhaps the balances of the economics are different for the more expensive DJI drones, but to me the spark seems very well priced for what it does. I wouldn't be building one myself if I wasn't at least as much interested in the fun of the project than the final result.

I suspect that it's a lot like building your own PC these days.

Around the end of last year I decided that I needed to replace a couple of very venerable DIY Windows boxes I had for gaming and other purposes. I started pricing out CPUs, graphics cards, motherboards, and so forth and I ended up just buying a big Alienware laptop. It fit into my environment better, was certainly fast enough, didn't cost much more, and was just less hassle.

I've spent many, many hours building PCs but I was sort of "been there done that" about doing it again.

I've been flying things for 30+ years, and the one thing I have learned in this hobby/field is that the cycle goes like this:

  1: Build -> 2: Fly -> 3: Crash (goto 1)

DJI are turning this into:

  1: Buy -> 2: Fly -> 3: Crash (goto 1)

All of these phases are ripe for disruption in my opinion. Maybe you'll find that building is way more fun than flying (crashing) - in which case, you can ship Step #2 off to someone else to deal with .. ;)

You piqued my interest. Now I wonder, how can we innovate in the space of #3? What disruptive potential awaits there?

https://flybrix.com/

    4: -> Give to someone else to repair...

3DR tried to compete with DJI in consumer space but failed.

By now, DJI has more than a thousand of employees and at least a decade of experience in this area.

.. and there are 10's of thousands of people out there who can just roll their own drones. Doesn't mean DJI isn't going to succeed - just that this is where they'll get their competition from.

I am trying to understand the point of your post or at least in what way it's relevant to measure the success of robotic startups.

neither iRobot or Kiva can be called startups any more. Then there is a company like Universal Robots which seems to be doing fine but hardly a startup anymore as it's went public and was bought.

There are plenty of robotics companies which are doing just fine as there are plenty of problems out there needing robotics and there is still some way to go until most organizations have renewed their "machine park" and thus in the market for more optimal solutions.

A startup is by definition almost always no profitable as it haven't found product market fit yet. Once it does that it become just another company or get bought by the big guys.

If you mean were is the Facebook or Google of robotics then sure but I don't think you can apply digital space thinking to the robotics space.

You could say that nVidia is a very successful company because of it's robotics support.

But again not sure I know how to interpret your post (not that there is anything wrong with it)

How do you tell what a successful robotics startup looks like at an early stage? Looking at robotics startups that succeeded tells you what to look for.

Not necessarily if those startups who become successful end up satisfying the market. So you never really know.

Stupid question:

Could it be that the lack of manufacturing in the US (when compared to China) is hurting the robotics industry?

US manufacturing output is at an all time high.[1] (Finally back from the 2008 crash.) Much manufacturing is automated, but outside automotive, doesn't need robots. Mostly special purpose machinery. A pick and place machine for assembling PC boards is not considered a robot, although it has much the same parts as an industrial robot arm, including a vision system.

[1] https://fred.stlouisfed.org/series/OUTMS

US manufacturing output has had a boom because we include natural gas and oil as manufacturing output, not just because of automation.

I would argue that US manufacturing has had a "boom" because no other option existed.

When you find yourself at the bottom of a deep hole the only way to go is up.

That's what happened with the crash of 2008. See the chart posted by Animats, here again for convenience:

https://fred.stlouisfed.org/series/OUTMS

As an aside, I love it when politicians take credit for naturally occurring phenomena like this. Industry go pummeled during that crash and every single business worked hard to claw their way out of the ditch. And then politicians proclaim "The sun went down and we made it come up every morning for the last six years!". You couldn't make this shit up.

We associate manufacturing with robotics because robotics has been very successful in factories. But growth in robotics is going to come from moving into spaces that aren't already full of robots. There are a number of companies that have moved into warehouses (the startup I work for among them) and probably you'll see shipment and retail stocking robots in production before too long.

what about Boston Dynamics, the military contractor. There's also Festo, Fanuc, Reis, I'm sure there are a ton of other robotics companies that we don't hear about because most robotics companies are marketing to very specific applications on the business side rather than the consumer side.

BD hasnt produced anything marketable still. Festo is an automation company and Fanuc is one of the old industrial robotics makers. Reis also just does industrial solutions.

None of those are startups besides BD.

Last I heard, Rethink has rebounded and doing very well. Their new robot has been selling extremely well.

As of January 2016, Rethink had received $115 million in investment and had sold only 900 robots.[1] They were selling the "Sawyer" robot by then, but no sales figures for that seem to be available. About 40% of Baxter sales were to research labs.

[1] https://robotenomics.com/2016/01/11/the-facts-about-co-bot-r...

Do you have a reference? I would be very (and pleasantly) surprised to hear this.

Baxter had two core features: its OS and its compliant joints, with the aim of it working next to humans both safely and easily reprogrammably.

They had to abandon the OS and switch to ROS which removed the point of being safe for use by naive people.

I don't - it's just what I heard through the grapevine from people working in the field.

I worked at Rethink (previously named Heartland Robotics) as an intern in its very early days (2009, there were ~10 employees). Exciting times.

Interesting tidbit around the founders:

> Boniske said it comes “back to the founders” and that, in her view, an ideal robotics founding team probably has three people, including the CEO who’s “going to be the visionary and drive the strategy of the overall company; this person is going to be doing the fundraising and recruiting. Then you’ve got the technologist who’s going to go and build out the engineering team and who ideally has a lot of robotics experience.”

The third piece — and it’s missing for a lot of teams, she said — is someone who can communicate the exact value proposition for the customer and speak their language.

Former colleague moved back to Iowa to work on an agricultural start-up. Given how prominently agriculture features in most drone pitch decks, I asked him what he thought of them.

"They're great! We check them out every time they come to town. Usually bring the kids, too."

"How do people like them?"

"They're nifty devices! Really futuristic."

"I mean when they use them."

"Use them? For what?"

"..."

"Haha, hell no. Those whippersnappers fly out here once a month at best. When I need something, I go to our local tractor vendor my family's done business with for generations."

> The third piece — and it’s missing for a lot of teams, she said — is someone who can communicate the exact value proposition for the customer and speak their language.

So, basically, someone doing sales and marketing. I'd think that'd be a CEO/non-technical cofounder job in a very early stage startup, but maybe the capital-intensive nature of scaling up a robotics/hardware startup means the CEO should be entirely focused on raising money and securing partnerships while someone else is entirely focused on product leadership. I wonder why the investor settled on that division of labor.

China-based service-oriented robotics startup founder here. This is an interesting but very US-centric perspective... over here in China, I don't think it necessarily holds. What would be different? In my interactions with western and Chinese investors thus far, I haven't seen much emphasis on team makeup... most queries related to business model.

In my experience all investors say team is really important. However, they rarely prove the team, ask much about their experience or background.

I think it's probably code for "has done a startup before" and "I kind of liked them".

Reply on news.ycombinator.com