Today is Orbotix day at Feld Thoughts. I’m sitting in the House of Blues in Chicago getting ready to watch the Excelerate Labs Demo Day practice pitches waiting for everyone to show up and listening to the sound check guys blast music and say “hey hey 1 2″ over and over again. I put my headphones on and listened to the new video by Sphero announcing and demonstrating their six new apps – Color Grab, Tag, Exile, Doodle Grub, a new version of Golf, and a new version of the core app which is massively upgraded.

Between the brain transplant and the six new apps, it’s a huge refresh for the existing Sphero customers. And, if you don’t have one, go buy a Sphero now.

It was the second week of TechStars and I was doing office hours with each of the 11 teams going through the Boulder 2010 program. I was sitting across the table from Adam Wilson and Ian Bernstein who each looked tired and dejected. In front of them were three slides.

I asked them what was wrong. They said they were having trouble deciding which of three different products to pursue. They’d had a dozen meetings with different mentors and were getting wildly conflicting data, which we refer to in TechStars as “mentor whiplash” and is a normal part of the first 30 days of TechStars for every team.

A few weeks earlier, Adam and Ian had their company Gearbox accepted into TechStars. They were hardware / software / robotics nerds and loved to tinker around. Their TechStars application and video had something to do with robots and their crazy desire to stay up all night hacking on them. When I’d last seen them a few weeks earlier, they were full of energy and life. Now they just looked defeated.

“Tell me about each of the ideas.” I asked. Adam started with the first one. “It’s a door lock controlled by your smartphone. The door lock market is really big.” said Adam in the most deadpan monotone voice I’ve ever heard.

“Ok – what’s the next one?” I asked. Ian mumbled something about the second slide on the table. I don’t even remember what he said.

“What about the third one?” Adam chimed in again, a little more animated this time. “It’s a robotic ball controlled by your smartphone.”

“Why are you having trouble deciding?” I asked. Adam kept going. “Some of the mentors like the door lock market but it seems like a really easy product to create and there are lots of door lock companies. And some don’t like it because it’s not defensible. No one really understands the second idea. And then there’s the ball – some mentors love it and others hate it.”

“Well,” I asked, “Is there one you love a lot more than the other two?” Adam and Ian looked at me quizzically, the same way my golden retrievers do when I ask them if they want to go for a walk. “Really, you just want to know which one we love the best?” asked Ian.

Adam jumped in, “Are you kidding. The ball. It’s a robotic ball you control with your smartphone.” Adam stood up with a gleam in his eye. “Brad, IT’S A ROBOTIC BALL YOU CONTROL WITH YOUR SMARTPHONE! HOW COOL IS THAT.”

At that moment, Sphero was born. And I knew that if Adam and Ian could make any progress over the next 10 weeks creating a robotic ball controlled by a smartphone, I wanted to invest in these two amazing guys. Their passion and obsession around the idea of a robotic ball you control with a smartphone was awesome.

Over the ensuing weeks they were regularly asked “what’s the market”, “who is the customer”, “how big is the market”, “isn’t it just a toy?”, and a bunch of other skeptical questions. As the weeks went by, they kept answering these sames questions with some variant of “we don’t really know but here are some ideas.” They kept working on the prototype and once they could drive something in a sort of a straight line, more eyes started lighting up and the “how big is the market questions” started to diminish.

Fundamentally, they didn’t know the answer, nor was it important at this stage. There are a zillion balls in the world and an endless set of applications for them; at this point Adam and Ian were pursuing the vision of a product that they were obsessed about. While some mentors and investors wanted to understand all the market and customer dynamics, others were able to see, or dream about, the enormous potential opportunities if the product could ever be created.

Three months later Foundry Group led the seed investment in Orbotix. And we are just as passionate about it today as Adam and Ian were way back when.

This morning we announced that we have just led a $3m investment in Modular Robotics. I’ve joined the board and am psyched to now be an investor in two Boulder-based robotics companies – ModRobotics and Orbotix. In the spirit of “show” vs. “tell” (which everyone in TechStars has driven into their brains over and over again) here’s a short three minute video that shows how ModRobotics first product – Cubelets – works.

Dan Primack at Fortune has a nice article up about ModRobotics titled ‘Robotics for kids’ start-up raises $3 million. For a little history of the rapid progression of the robotics movement in Boulder, take a look at my post titled Boulder is for Robots. And – if you are in Boulder and interested in robots, join the Boulder is for Robots Meetup Group.

My partners at Foundry Group and I are psyched to be working with Eric and team on creating the next generation robot construction platform. I believe the machines have already taken over and are just patiently waiting for us to catch up with them. As part of this, every kid should learn how to program, create things, and work with robots. ModRobotics plans to be an integral part of that.

Oblong is one of the most amazing companies I’ve ever had the honor to be involved with. We invested in them in 2007 when they were four people building on the incredible research and ideas of John Underkoffler. Today they are almost 100 people strong, have shipped an awesome set of products, and are on a path to fundamentally change the way we interact with computers.

If you are in the Washington DC area and want to see a live demo, email me.

Today one of our portfolio companies, LinkSmart, has come out of stealth mode and unveiled its product, Total Link Management and $4.7M in financing led by Foundry Group. My partner Seth Levine sits on the board and wrote about his view of LinkSmart. The investment was an obvious one for us as it fits squarely in our Adhesive theme given the deep relationships LinkSmart has with big media and web platforms as a provider of analytics and traffic management software.

Even though the Internet and web has evolved since the mid-1990s, the most prolific web properties on the Internet all generally follow the same process: advertiser relationships are created with a publisher. This publisher then runs a never-ending fire drill with their sales, SEO and yield management teams to fulfill those obligations with their advertisers. More ad units are inserted on pages, marginally better eCPMs negotiated for, but yet the profit margins continually come under pressure as publishers scramble to generate even more traffic to support the advertising expectations.

LinkSmart has built a system that fixes part of this challenge. The technology LinkSmart’s founder Pete Sheinbaum and his team have assembled is analogous to many of the solutions we’ve seen and invested in for search engine optimization (SEO) such as SEOMoz. However, while SEO directs visitors from the open web — via search engine results — to a publisher’s site, LinkSmart takes those same readers and provides the publisher team with the means to use hyperlinks to more deeply engage the visitors, keep them on site longer, or send them to the properties the publisher wants them to visit.

If it costs a dollar to bring a reader to a site and they bail after the first page, that’s a dollar not optimized or even wasted. If that same dollar is spent and LinkSmart moves the reader three pages deeper on the site, triggering more revenue or visits to affiliate or partner site, the cost per acquisition has dropped dramatically and in many cases will improve the RPM geometrically.

Beyond the link optimization, there’s an enormous amount of valuable data generated by LinkSmart’s system. The data answers questions such as what are readers clicking on? What is their path while on site? What is the optimal number of links 100 words of written content should have to guarantee highest click-engagement? How many readers are leaving your site in a single click and where are they going? LinkSmart helps a publisher look deeply at that data to surface many things including the real-time intent of a reader when on-site.

We are psyched about what LinkSmart is launching today. If you are a publisher, check it out — it’s just the tip of the iceberg.