Blog | Robotiq

5 Internet Attitudes to Apply to Robotics

Written by Samuel Bouchard | Nov 03, 2012 4:37 AM

Last week at Robobusiness, a panel of investors were debating on the "State of the Robotics Industry - Investment Trends". One of the panellist was Dimitry Grishin who launched Grishin Robotics, his private investment fund dedicated to the personal robotics sector. Although he is from Moscow, Dimitry looked like any Silicon Valley Internet millionaire speaking on a stage, chilling on his sofa between the 2 other panellists. When he started talking about where robotics should go with his strong Russian accent, cutting words and rolling r's, you could feel a boiling passion for robotics and for not accepting the status quo.

The successful Internet entrepreneur who had studied robotics at Moscow State University was arguing that robotics should "use more Internet attitude". This in an interesting thought that I've tried to apply at Robotiq as I also come from the Internet world. (I was involved for several years in the largest Canadian private sale real estate website before starting and investing in Robotiq.) Grishin was sharing the stage with Rodney Brooks (iRobot, Rethink) and Jan Westerhues (Bosch Venture Capital). The three of them agreed that there are two main differences between robotics and Internet businesses:

  1. Robots take more time, more resources and a more diverse mix of resources to develop. Making a robot means physically building mechanics, electronics and embedding software. In the Internet these days, you don't even need infrastructure, two programmers plus an Internet connection and you're started.
  2. In robotics, you have to physically ship the product. You're not only sending data over the network as with Internet-based companies.

Still, they also agreed that robotics startups could benefit from having an Internet culture. Here are a few ideas that Grishin brought to the discussion.

  1. Use agile development

    Agile development means having a high customer sampling rate to control your product evolution: showing the product early, often and iterate it until you have something people are willing to buy. With the Internet, closing that loop is easy. You have web-based users that give you comments. At some point you charge them money. Everything can be easily tracked and analyzed. If something needs to be modified, your software can evolve quickly as it is only code, no hardware is involved.

    Why do that in robotics? Because development time and cost are even more important. So if you take 1 year to develop your product with no customer feedback, the product-market fit will be poor and you won't sell. Then you have to go to another long iteration and you end up loosing time, money and opportunities.

    How to do that in robotics? Talk to customers right from the start and keep them involved. Use concepts, drawings, rapid prototyping, videos to get feedback during development. If you need to build a very complex system, try to break it into a roadmap in which components can be validated along the way.

  2. Use building blocks

    Building a web site today is very efficient because there are so many well interfaced building blocks like databases, libraries and APIs available. So people are really focusing on the product integration and features, not as much on the underlying technological parts.

    Doing that in robotics is easier said than done. The main challenge here is the lack of mechanical and electrical standards. It is difficult to choose THE building block that will work with everything. Still, we see some people applying this concept. One example is Double Robotics, in which Grishin invested, they use the iPad as a building block for their telepresence robot.

  3. Use e-commerce

    True, you can't ship a robot through fiber optics. But you can at least cut out the middleman and be more cost effective by efficiently using e-commerce. And by using inbound marketing best practices, you can have very efficient web based lead-generation too.
     
  4. Focus on software 

    Even if robotics has an important hardware component, Grishin suggested having a good dose of software in your product. Why? Because software does not add to your bill of material, but can add tremendous value to your product, leading to better margin, more differentiation and a higher barrier to entry. Sell a simple mechanical widget and soon you'll have competitors fighting with you on price.
     
  5. Embrace connectivity

    The value of many Internet based companies is not what they are but how they could connect a huge number of users. Think of the fax machine. The first one that was sold was pretty useless. When there were 2, they could send messages among them. When there were millions, then the fax became really useful. Connecting your devices can again increase its value without adding cost. 

It is always interesting to look at new approaches and innovations for products, but I am also a big fan of the basics. And whether you are web-based or hacking robotics hardware, there are things that will always be needed and the panellists agreed: Focus on problems people have, and problems they are willing to pay to solve.