Case Studies

The company we were working for had built a prototype based on using the first-generation Kinect camera for depth sensing. They had removed the plastic covers and other extraneous parts (microphones, geartrain, etc) and cut the sides of the sheetmetal frame in order to fit the device into their product. Creative idea for a quick proof-of-concept prototype! 

When we were brought in to help this group to scale the product, however, we quickly identified that this depth camera design would not scale for production. The Kinect sensor uses structured IR light and an infrared camera together with a proprietary chip from PrimeSense to produce the depth image. The camera and IR projector must be precisely calibrated in order to produce an accurate depth map; taking the product apart and cutting it up was more than sufficient to drastically change that calibration. So while the POC was sufficient for a quick demo, it was not in fact giving accurate data!

We helped this team to design a new optical bench and new PCBAs and FPCs to make the connections in the new system, worked with a manufacturing test and calibration organization to build a calibration process based on the spec from PrimeSense, built an assembly flow and fixtures, established a supply chain to purchase the cameras and IR projector, and handed all of this off to a contract manufacturer to produce the sub-assembly. 

This is a great example of the way we work, looking not just at the task at hand (i.e. help us to build more of this), but also at the broader picture of meeting the team and organizational product goals and specifications. Our expertise in opto-mechanical design and manufacturing uniquely positions us to quickly understand the risk/reward tradeoffs and help organizations make the right sized-choice for their product. 

Concurrent was brought in to help obtain more of an optical sub-assembly being built by a third-party vendor. There were some issues with initial prototypes from said vendor, but the internal team of our customer were not familiar with optics so they could not really ascertain what needed to happen to obtain more of the sub-assembly. As far as they knew the design was complete, they just needed to buy more for additional prototyping needs and plan for scaling to production volumes.

After some initial assessment by our team, it became clear that the design of the sub-assembly was inherently not scalable, because it involved purchasing a development kit from an LCOS vendor and re-using its components, and that dev-kit was no longer in production. The most volume that could be made using the current design was about 15-20 units. We simultaneously kicked off working with the existing supplier to obtain those parts and build the current design, while also doing a broader vendor RFP with multiple potential partners to re-design the optics of the module. 

The new design was not only able to solve the supply chain constraints, but also able to optimize for the desired specifications - reducing weight by 25%, and improving critical optical performance metrics by 10%. Now that the client owns the entire design, they can create production tools and save 75% on the module cost as well!