As explained in our previous update, the last run of Luminodes were not manufactured according to our exact specifications nor using processes suitable for final manufacturing. Despite these issues we were still able to do some additional testing on some of those units and located a design issue related to our sensitive internal safety fuses blowing prematurely due to the inrush current during startup. To eliminate the problem we replaced the fast blow pico fuses we were using with even lower power, but slow blow fuses and a power resistor to also slow the inrush current. To make this fix we had to modify the circuit board design as well as one of the plastic enclosure parts. This of course introduced an additional unexpected delay.
Fortunately the UL test facility has already approved these minor design changes and we have now begun manufacturing these new Luminodes so that we can resume UL testing soon after receiving the new units we expect during the first week in June. Once we verify these new Luminodes and submit them for the physical UL tests we will provide a new updated ship schedule.
As always thank you from all of us at Think Automatic for your ongoing patience and support!
For those that are curious for a peek at some of our new test processes used to help isolate potential issues during manufacturing of these early small runs here is a video we provided to our manufacturer showing them how to use the test module (warning – it’s not glamorous I built the test node out of a Freescale Sensor Node reference design module. It took some coding finesse but I was able to turn the Freescale Sensor Node into an ISSP programmer so that it could program the Cypress PSoC chip on the Luminode mainboard and then verify the various connections such as the I2C pins, the buttons, the tri-color LED, and the relay that we use to satisfy the UL required air-gap safety feature.
And for those curious about our FCC testing processes for the RazBee (which we expect to have FCC approved soon) here is a video sent to the FCC test facility to help them put the RazBee into the test modes they need to in order to certify it. Enjoy! It’s gripping! I mean, kind of mundane…
Last week we received twelve new Luminodes with the required UL design modifications in place for testing. Although the new mechanical parts fit and work great, the electronics were not manufactured to our specifications or the necessary quality to complete UL testing. We are defining new processes to make sure when we begin reproducing the electronics later this week they will be at a production level of quality sufficient not only for UL test purposes but also for final manufacturing.
Obviously this minor setback means an additional delay from our previously planned ship date of the end of May. Once the re-manufacturing of the units for UL testing begins we will provide an updated timeline for shipping Luminodes.
As always, thank you for your patience and support.
Check out our new video update to see Luminodes and our new mobile web UI in action! Also get a closer look at the Raspberry Pi based RazBee with RazThing wireless sensors programmable not only to work as door and window sensors through our scene learning platform but also temperature, movement/acceleration sensors and a host of other types of sensors just waiting to be built using RazThings!
Thank you for all your support from all of us at Think Automatic and the Luminode project!
UL approved our design last week and our new test units are in manufacturing right now! We expect to receive those new parts and assemblies on or around April 17th. We will then assemble and program those units and ship them back to UL for the physical tests which if all goes well will take 2 weeks from then. Once we have full UL approval we will immediately begin manufacturing the final units which will take around 3 weeks.
Long story short we are now aiming for late May to be shipping out final Luminodes.
Now that we are getting closer to shipping I will try to do an update each week, even if just to say still on track.
The video that I mentioned in the last update we had to reschedule to shoot on this Friday. So the next update should have some video too!
We have been very busy working to get the Luminodes into all our backers’ hands as soon as possible! We have completed the design modifications to the Luminode which we believe addresses all of the issues we encountered with the UL review. We are in the process of modifying the manufacturing tools and we expect to run new parts very soon so that UL testing can continue. We are on a very accelerated schedule but still hoping to be in production by next month. As usual thank you for your continued patience!
In addition we have made solid strides on our machine learning platform making it freakishly easy to incorporate machine learning into virtually any device that can talk to a computer. It is straight forward to integrate everything from the SmartThings puck (we’re backers ourselves) to the Ubi (who we are collaborating with) to RazBee devices (which we are closely involved with) to Twine to a Nest thermostat to a HeatMeter to your own projects and sensors!
Unlike many other types of devices we believe applying machine learning to lighting control takes a particularly special touch, which is the reason why we developed the Luminode in the first place. This means that the Luminodes are better able to showcase the rich capabilities of our platform more than any other dimmer on the market!
Once integrated (did I mention it’s easy to integrate with?) all your devices become part of a platform for a new breed of user experiences that are a super set of the conventional UI’s you typically only see on phones and the web. Although we will provide to you (for free) our own phone and web UI’s, it is also very easy to integrate our system into your own systems and UI’s if you know how to program against RESTful services.
We are a tiny operation working to do something big and we’re getting close! All the pieces are falling into place and we plan to shoot a new demo video by next week showcasing some of the cutting edge capabilities I just described.
As DIY’s and makers ourselves we know what it is like to build your own system and what we are doing is to make that easy to the point of being nearly automatic!
Thank you from all of us at Think Automatic! The finish line is in sight!
PS – We want to give a special shout out to the HeatMeter folks. Like us they’re engineers more than marketing and business folks and their project has only a few days left… Please check them out!
The good news is that the Luminode now has some nice new packaging and a solid plan to reach production!
The less than good news is that UL testing began behind schedule and UL is just now requiring us to make some minor but important design modifications to satisfy UL standards. Specifically we need to increase the spacing between some components and circuits on the primary circuit board.
In order to make enough space on the board for these changes and to minimize the risk of additional UL concerns, we have made the decision to remove the on-board current sensing circuit. Although we regret having to remove that feature, we believe it is the right move to make in order to get the Luminode out into the world in a reasonable time-frame from now. Rest assured that the Luminode will still enable energy savings through minimizing when lights are on unnecessarily even without the current sensing circuit.
We expect to complete the necessary design modifications by early next week, have the changes reviewed by UL, modify the tooling, run new parts, and resume UL testing, all by mid March. Although this is an aggressive schedule we believe it is achievable and we hope to now ship Luminodes by the end of April. We understand this is a full quarter behind or original target ship date of January, and we are truly grateful for you continuing patience and support. UL approval is an important safety regulatory certification that we must comply with and we are dedicated to ensuring that you, our earliest supporters and customers are not only delighted by your initial experience with the Luminode but also have the piece of mind that it is thoroughly safety tested and approved.
Please accept deep gratitude from all of us at Think Automatic and the Luminode project!
Last night the man who literally wrote the book on crowdfunding moderated a diverse and lively panel on the topic for the MITEF’s venture lab—and though I sat in the back of the packed room at Union Square, this passenger had a front row seat in the crazy ride that is a Kickstarter campaign. Full disclosure: my husband Steve Harris , the founder and CEO of Think Automatic, sat on the panel, along with Harebrained Scheme’s Mitch Gitelman and Uber Entertainment’s Bob Berry. Color me biased but I thought the group had some awesome take-aways, so I thought I’d list a few here:
Fund a project, not your company: Though the panelists’ Kickstarter pledges ranged from $36K to over $2M, one theme was consistent. Kickstarter is not a means to fully fund your venture, but can be a great way to “get paid to advertise and interact with customers” as well as obtain pre-order sales for a specific, tangible milestone in your company’s development. In the case of Think Automatic, Steve’s been contributing his time (and savings account) to the business on and off for over seven years. The Kickstarter campaign was a great way to pay for tooling, and start manufacturing the smart lightswitch for his home automation venture. Steve’s backers were motivated to get the company over this exciting hurdle, much like the Shadowrun and Planetary Annihilation fans knew there would be kick-ass games at the end of the Kickstarter rainbow. The point: carve out funding goals (and stretch goals) with tangible outputs so your backers can see their impact in specific and satisfying ways.
Money can’t buy it: earned media is priceless I’d love to tell you that Steve’s paid advertisement on GeekWire drove the lion’s share of the traffic to his Kickstarter video, but in fact it was this killer article on Engadget. Kickstarter both has its own PR engine (through featured campaigns) and can serve as a great reason to reach out to the media. As always, know your audience, find the best media conduit to the kind of makers, gamers, or other folks who may be interested in your campaign, and sell your lights out with an interesting hook.
It’s all about the video: Whether you spend $30k on a professionally produced video (as did Bob and his team), or go low-budget hilarious (Penny Arcade style), a great video is the lynchpin of a successful campaign. Above all, authenticity is key: the style should resonate with your brand and your audience. Charmingly goofy nostalgia worked wonders with the avid lovers of Shadowrun, and the geeky maker-esque Twine video from MIT grads helped yield 140X their fundraising goals.
Ya gotta want it: If you embark on the Kickstarter adventure, you’ll have time for little else whether you’re a two-person team or a 20 person company with eager interns. Be ready to commit yourself to daily or weekly updates, new videos, and responding to seemingly endless comments and questions, on Kickstarter itself and via their email engine. Beware of hidden costs such as picking, packing and shipping the shwag you promise to your backers, and consider the relative simplicity of digital rewards. That said– to a person, the panelists agreed these were good problems to have to handle.
So if Kickstarter is the right path for you, buckle up and enjoy the ride!
We were hoping that we wouldn’t be yet another Kickstarter project that slipped their target ship date, but at this point it has become unavoidable. Please accept our sincere apologies and continuing gratitude.
Our trial run of 50 units was delayed slightly due to a minor manufacturing glitch and we now expect to receive those test units mid next week. UL testing has also taken longer than expected but we are still hoping to receive approval within the next two weeks.
We have also initiated both FCC testing and manufacturing of the RazBee boards enabling Luminodes to talk directly with the Linux based Raspberry Pi’s complete with our machine learning integration system which we plan to ship in place of the USB dongle at no extra cost (or in addition to for those who wish to purchase one separately).
It is also becoming apparent based on our manufacturing costs that our target retail price of $65 per Luminode ($60 in quantity) is not going to be cost effective for our early production runs. We are however going to continue to offer discounted pre-orders for our very first run targeted for February as a special thank you to our earliest adopters!
To pre-order additional Luminodes or starter kits at Kickstarter prices please visit our online store here: http://blog.thinkautomatic.com/shop/
Thank you for your continued patience and support from all of us at Think Automatic!
It’s been a little while since our last update and we’d like to give everyone a quick update to let you know that everything is still on track both for the Luminode and now for our Raspberry Pi based learning integration system.
We are expecting to receive our first test production run of 50 Luminodes by the end of next week and we hope to hear word on UL approval shortly thereafter. Once we have UL signoff and have verified the quality on the 50 Luminode test run we will initiate our first full production run! Assuming all goes well we will be shipping Kickstarter rewards and other pre-orders by the end of January!
Not only are we shipping the Luminode we are also pursuing FCC approval for the RazBee radio board! This turns the Raspberry Pi into a Luminode gateway and learning based system integration device! It is also an open platform that can be reprogrammed with ZigBee or other protocols.
If all goes well we hope to offer to Kickstarter backers whose pledges include the USB dongle the option of receiving a full RazBee kit including Raspberry Pi in place of the dongle for no extra cost to you! Stay tuned for more details!
Thanks from all of us at Think Automatic and the Luminode project!
We have passed FCC!! And UL testing is scheduled to complete in mid December!
We are also initiating a test run of 50 Luminodes to make sure our manufacturing facility is prepared to make our initial production run once we get UL approval.
In addition our electronics hardware designer Steve Dame has launched his own side project to compliment the Luminode or your own Raspberry Pi based wireless project! It is a small programmable daughter board for the Raspberry Pi called the RazBee. It is based on the same Freescale MC1322x chip as the Luminode which can be loaded with our Luminode coordinator firmware or your own firmware project! He plans to launch his own Kickstarter project to bring it into mass production. You can check out his website here.
Meanwhile we have successfully ported the Think Automatic home automation learning algorithm to the Raspberry Pi with the RazBee daughter board! Here it is in action:
If all goes well, by the time we ship the Luminodes in January we will be able to offer the option of a full Raspberry Pi + RazBee pre-loaded with the Think Automatic learning system as an alternative to the the USB dongle at little or no extra cost! We’ll keep you posted as we progress…
I built the test node out of a Freescale Sensor Node reference design module. It took some coding finesse but I was able to turn the Freescale Sensor Node into an ISSP programmer so that it could program the Cypress PSoC chip on the Luminode mainboard and then verify the various connections such as the I2C pins, the buttons, the tri-color LED, and the relay that we use to satisfy the UL required air-gap safety feature.
I mean, kind of mundane…
