HardTech Blog

GigaOm posted an article about how hardware startups are a completely different game, there’s also a little bit of back-channel on Hacker News. Worth a quick read.

http://gigaom.com/2012/02/24/building-stuff

I thought I’d repost this article. Basically, they look at MakerBot and Adafruit as very successful open source hardware businesses, and look whether or not investors are amicable. Brad Feld weighs in saying he thinks MakerBot is awesome, but I didn’t hear about whether or not the open source was part of it. The quote was more about MakerBot “being the Apple of HP of this space” which sounds like standard VC pro-profit talk - not as much as open source adds value to society, or at least to companies. We’ll see where this goes!

http://www.businessweek.com/magazine/the-diy-maker-movement-meets-the-vcs-02162012.html

Just ran across a cool Kickstarter, called Twine, that allows non-programmers the ability to connect sensors and other input devices directly to their computer / internet, and then they can use a verbal interface “program” the behavior. Think of it like a Siri for sensor programming - or if you haven’t seen the pure-web version of IFTTT (that stands for “If this, then this”), it’s similarly cool.

Internally it has a temperature sensor and accelerometer, and externally they’ve got magnetic switches (think: door alarm), moisture sensors, and the obligatory breakout board. 

The image they give on Kickstarter isn’t very inspiring… Looks more like fancy soap in the picture, but if you watch the video you can tell it’s more than a hunk of green plastic. I’ll give them a partial pass though, because their web interface, called “Spool” (cute) is a lot snazzier and straight forward to use.

I really like the concept and think it’s moving in the right direction. I have to say though - the sensor choice is a bit underwhelming. Seems more like a weather measuring device than an easy link to the internet. What am I going to say? “When temperature goes below 65 - Tweet: ‘Getting a little chilly in here’” Or maybe the accelerometer can be used to detect if the user is pounding on their desk? I don’t get what they’re thinking with these choices. 

I have no idea of their future intentions, but I think it’s safe to say that the future will see many interfaces like Twine, Siri, and IFTTT that let non-coders write their own scripts - without even knowing it. I’m excited to see where this goes, and hope that they put their funding to good use in getting this shipped.

The proliferation of maker culture and especially cheap 3D printers has driven the demand for easy content creation tools. The community mostly used either professional CAD (often torrented), or they used Google SketchUp or some form of open source tool like Blender or OpenSCAD. None of these are very good at the full job, and only OpenSCAD could export a clean STL file (for 3D printing) - but you had to be a programmer to get it to work. Blender and SketchUp were easier to use, but not really designed for rigorously representing physical parts.

This physical-stuff trend, along with the advent of 3D-friendly internet standards like WebGL, is generating some new startups on the block that are trying to give these tools to the masses, so I wanted to go over a few of the ones I’ve stumbled across:

First, is TinkerCAD. It was only about a year ago that Kai Backman let me play with the early demo of TinkerCAD, and now they’ve gone and raised $1M from True Ventures! They’ve definitely made a clean and very intuitive system that that’a easy enough to pick-up and use. The interface is responsive, as it’s done client side, but the heavy-lifting takes place on their servers with a little latency, but pretty useable. 

Here’s a screenshot:

Another interesting point where they’ve made progress is their STL export. Early on, it was freaking HUGE, but now they’ve slimmed down the polys. Check out this screenshot of the mesh in NetFabb: TinkerCAD Part Mesh 

Next up, I just found out about ShapeSmith, and I’m really impressed with what they’ve done - best of all it’s even open source. It’s a very similar tool to TinkerCAD - a little more dev-oriented - not as quirky / fun looking as TinkerCAD, but it’s got some neat features. My favorite right now is the normal-CAD-like operation menus where you can enter real dimensions.

Another interesting aspect they’ve got is that the back-end is using OpenCASCADE - which was built for more hardcore things than just 3D printing. Unlike the STL/mesh representation that TinkerCAD has, ShapeSmith used a “B-Rep”, or Boundary Representation, which is a more rigorous and modern way of making CAD files that allows you to edit it later - unlike meshes which are pretty much fixed. I’m interested to see how they progress.

I literally just heard about Sunglass.io over the weekend, and I’m not sure exactly what they’re trying to do. The FAQ says they’re trying to build a collaborative online CAD system - pretty ambitious. So far, they’ve got a nice smooth front-end that apparently takes OBJs - so they’re meshing it, but in color. Better than the raw STL meshes - so it’s a start. It doesn’t have all the functionality, but I’m looking forward to seeing what’s up in the future.

I’m not sure I fully buy that people beyond beginning want simple CAD software in the cloud just yet. The major issue we encountered at CloudFab was the internet pipe wasn’t big enough to get all the polys uploaded quickly - regardless of WebGL rendering them. It’s a problem is people can upload a lot of stuff, or design complex stuff, that really bogs it down, but I suspect for maker-CAD it’s probably acceptable. 

For what the major CAD guys think, check out Develop3D’s great post on Cloud CAD. The gist is that everyone is excited for what large-scale cloud simulation can do - which is a logical step for them to integrate into desktop CAD for cheap, rapid high-end analysis. However, all of them expressed concerns about going to the cloud for the design tools themselves - mainly for business model reasons, but I think they’re partially right about the look/feel still being less than professional-grade. I’m excited to see where all of this ends up.

It’s always cool when companies with a broad view of the market share what they’re seeing. OkTrends is the blog of OkCupid, and as sappy as you might think that could be, it delivers on the data! I dare you to have a look without going back.

Well, OctoPart in the past few months has been on a data kick of its own that I hope will continue. Two of these posts are:

• “The Far Limits to Datacenter Compute Efficiency” - which talks about what the real metric we should be talking about when referring to Moore’s Law.
• “Volume Pricebreaks: Double-Edged Sword for Smaller Organizations?”

The second one is extremely pertinent to HardTech, as I hear often about people struggling to find components and at what volume to buy. They have an awesome chart showing discounts per component type over many vendors - just beautiful… But also very instructive for those of you wondering what “volume” actually means on some of this. Well worth the read.

The Atlantic online has a cool write-up about different hardware companies that came out of TechShop: 7 Successful Products to Emerge from San Francisco Techshop

Quick Link List:

1. C-Loop
2. Square
3. DoDo Case
4. Embrace Infant Warmer
5. Lightning Motorcycle
6. Clustered Systems
7. Karat Stix

Something is rotten in the state of startups… The promise of technology, to further humanity by solving the real problems of our day, is in decay. We’ve gone from solving real problems to increasingly vapid me-too startups trying to make a quick buck on the flip. What’s gone wrong? Aren’t we seeing a massive amount of innovation today!?!

So yea, the world of the web has been unlocked so, for the cost of living, you too can build a web service and maybe a business along with it. The landrush is now in full swing to apply these new techniques to the frontiers of markets henceforth unseen.

We all know the keys to our new found superpowers:

• Open Source Software - We stand on giants’ shoulders - free of charge!
• Cheap / Scalable Hosting - No longer a fixed cost - often free.
• Ubiquitous Broadband - Yea, dial-up sucked…
• Community - Startups are a “thing” with advocates and meetups around the globe.
• Proven Techniques - Blank, Ries, Fried, etc. all helped show us nerds the way to salvation in how to really build a business, and we saw people just like us succeeding.
• Ample Investment - Those who cashed out before are now making funds and investments into others on the same journey. Investors with real experience and money - priceless.

All this though, and where have we taken this train? Daily deals, digital coupons, social “gaming”, fart apps, etc. Having moved to San Francisco a few months ago, I could fund my next startup with a dime from every Social-Mobile-GameMechanized-with-Deals startup I’ve run into. Please refer to this public service announcement for more details. 

This isn’t to say that the internet game is done, far from it, the cards are far from down, but the web has certainly matured greatly - which means less low-hanging fruit for us to work with. What I want to know is: what’s the next land rush in technology?

I don’t claim to know this, but I think we’re going to go through something similar to what we did with the web: lots of hard problems to solve that establish our superpowers and then a landrush of people using these powers to fill the void. In search of solving these hard, more-than-software problems, I’m starting this blog to look into the growing trend of startups that have a product based on making physical things.

There’s no good term for them, so I thought I’d make my own: HardTech - blending the elements of the difficult and the physical to form companies that solve real problems in the real world. This is the first page of my journey to explore the coming revolution, and I hope you’ll join me along the way.