Small steps. Big impact.

THE FITBIT BLOG

Fitbit production prototypes are created

| 40 Comments

Last week, we went to Singapore to watch our manufacturer create production prototypes that they will use to uncover any remaining issues and to tweak the design before the Fitbit is sent off to mass production:

 

Observing the assembly process Looking dapper

 

Here are some pictures of the production prototypes. You can see the OLED display shining through the translucent plastic case. Ignore the temporary scotch tape :). It’s there so we can pop the case on and off easily. The case color is also a bit reddish. We could not get the right color plastic in time for these prototypes, but the final version will be charcoal grey.

 

It's alive!  Four in a row

 

So where do we go from here? Typically, products are initially designed in some kind of 3D CAD (Computer Assisted Design) tool, such as Pro-Engineer or SolidWorks. Here is the Fitbit base station as a CAD rendering:

CAD rendering of the Fitbit base station

However, you need to build physical prototypes to understand the assembly process and to uncover any problems, either functional or aesthetic, that were not apparent in the 3D file.

Here are some examples of what we uncovered:

1. Assembling the first Fitbit was long. It took our manufacturer’s engineer 30 minutes to assemble. This is obviously not acceptable for a mass volume consumer product. The engineering team then came up with design changes to the device that will greatly accelerate the assembly process. For instance, it was really difficult to stick one of the circuit boards into the case, so the mounting mechanism was redesigned slightly.

2. The button press is way too stiff. We made some design changes and also adjusted the softness of the rubber of the button.

3. The charging contact mechanism was unreliable, mainly because the contacts would bend when we tried to insert them during the assembly process. We are trying to solve this by switching from phospher bronze to beryllium copper (which is stiffer) and by increasing the contact thickness. There are also some changes to the plastic housing that resulted from this. I’m personally particularly sensitive to this issue of contact reliability because I owned a first generation Logitech Harmony universal remote that was extremely difficult to charge. It drove me crazy because if I didn’t place the remote just right on its charging cradle, the remote would reboot!

In any case, there’s nothing unusual about problems like this occuring during prototypes. There’s always a gap between theory (CAD files) and reality (the physical prototypes).

At this point, it will probably take us another week or so to fix these problems and generate a few more runs of prototypes. Once we feel comfortable with the overall design, then we will move onto the process of creating the plastic molds for the housing, which will take about 4-5 weeks.

Also, remember to check out our photos on Flickr.

Happy holidays, everyone!