Design iterations

[Guest post by Jake #2] After completing a few more tests on the temperature simulations it was time for Phil and I to take another look at the design of the lightbulb.

Ben and Nathan (our designers) had flown in to Hong Kong the night before, crossed the border and made it to our Shenzhen hotel around 2am. We had already arranged an 8:30am pick up for the next day so on a few hours sleep they slammed down some breakfast and we all jumped in the van and headed to the factory.

Design after design was thrown up on the screen. We would all have a look, add our thoughts, make some adjustments then hand it off to Chen to run the temperature simulation. With each variation the bulb got cooler and the design got more refined, the foundation we had been working on for months was now all coming together.

One thing we really learned to appreciate was that it can be quite daunting to see how much designs can change, and adapt when you’re developing a factory prototype!

Prior to the trip I think we had a presumption that our initial prototypes, and the CAD designs for our desired configurations would be sufficient, but we’ve ended up iterating upon this design many times to strike the perfect balance between thermal management and lumen output, which is perhaps the most challenging aspect of all, but one we’ve now solved.

In lamens terms, this basically means as a lightbulb manufacturer you want to strike the balance between having a bulb that has great brightness (~900+ lumens, ~equivalent to 75w) and not having the bulb run too hot because LED lifespan is diminished if a bulb is “running hot”.

Thermal Management

Snapshot of thermal testing, I’ll see if I can get permission to post the full set of slides which includes four iterations we tested and have since improved upon.

Not only that, but once the CAD’s and the thermal testing have been done, you then have to make the bulb look sexy because it needs to be a physical representation of the love that’s gone into it.

It was great having Ben and Nathan both here with us in the factory because when you’re designing with thermal management in mind you don’t always come-up with the best looking design first go. For example, here’s one design iteration that we experimented with and had a cast made, not particular sexy right?


Render of a previous design iteration we toyed with.


Solid works prototype version of the design above.

Then Ben and Nathan lovingly take what we give them, use the same set of constraints, but alter the shape and design to turn it into something much closer to what the reinvention of the lightbulb should look like.

We’ve included this for illustrations purposes only, because these are certainly not our final designs, but you can see the iterative process we go through to get to the final result and bring LIFX from our initial prototypes into a consumer ready product.

Because design is such a critical aspect from a functional and aesthetic point of view, we’ve spent additional time on this than perhaps we’d intended. It has slowed us down a little, but it’s worth spending the time to get this right, a design that is great eventually, trumps a design that is OK permanently. It’s a philosophy we believe will deliver the best product on the market.

After this process we reached the weekend and Saturday was a catch up day. We had day to day tasks that had taken a back seat for the week whilst we focused on the bulb. Emails, contracts, supplier agreements and updating the team back home filled most of the day. With a few hours spare we headed into the city to do some awesome China style shopping.

Being involved in a start up I always expected there would be moments where I would find myself saying “well this was unexpected”. Standing in a market in shenzhen with my CEO negotiating a bulk purchase of underwear was certainly one of these:)

Sunday we had planned to take as a rest day off to recoup a little but with everything that had happened during the week we ended up spending most of the day sitting on the balcony talking LIFX strategy and bulb design options.

Monday morning. The guys at the factory were busy stamping out five revised factory CNC rapid prototype bulbs for us. Marc decided to head in for the day while Phil and I camped out in the hotel room trawling through more emails and “general business”. By late Monday afternoon we checked in with the guys at the factory and arranged for the next days visit. Marc arrived back to the hotel about 10pm then came to use my microwave to reheat some dinner. (for some reason my room was the only one that had a microwave??) We discussed the events of the day, and our plans over Chinese New Year, and how we feel we’re progressing with the project overall.

Ben and Nathan doing what they do best.

Nathan, working his magic on the various iterations of the designs we’ve been developing.

It’s an exciting time, and we’re certainly at the business end, with the hardware coming together in a rapid iteration fashion, and the software being developed in parallel in Melbourne. We’re getting closer each day and the rush it brings is phenomenal. We’re learning too, the importance of spending a little extra time to get things right. That being said, on schedule we inserted our control boards into an earlier design and have a working factory prototypes, but the extra time spent on design is something that will yield a product that we’re truly proud of. We’ll write a separate post about this and show the pictures etc because there’s a lot of detail to go into.

Speaking of control boards, some of you wanted a better idea of how tiny these are, so below is one that shows some scale. We’ve cropped the image just because we can’t show our full chipset and configuration out in the public domain just yet, but hopefully it gives you the idea 🙂

LIFX Control Board ;


Underwear shopping with Phil. Certainly not my favourite way to spend the only weekend off 😉 


  1. Great post guys. It really is interesting hearing how the process unfolds. Keep up the good work and the posts.

    1. Sure thing Greg. Sit tight for now, when the bulbs are ready you’ll be sent an email directing you to your unique delivery page. From there you’ll be able to update your delivery address before they go out 🙂 We learned this from ScanBox that the most effective way to take address changes is as close to delivery date as possible, and add that extra step so you’ve got control of the delivery address.

  2. Are you including consumer tests/feedback before finalising design and features? (The Lean Startup – Eric Ries etc.) – or are the Kickstarter first orders from your “tribe” actually the test? Thanks for the transparency of your process and the interesting updates – wish you every success and that LIFX becomes a model innovation method!

  3. I prefer the shape of the lightbulb in your sketches and in your logo. This iteration of the LIFX lightbulb looks like the Phillips Hue version. Too directional with a small flat light surface and too ordinary. I want the lightbulb to light up inside a translucent light fixture, like a shade lamp or glass globe, in an even manner.

  4. It’s really very interesting! I’d like to see and meet this lamp as soon as possible! Shape is still on designing or fixed already?!

Comments are closed.