-          Developed a process for prototyping different shoe integrations internally using cast urethane and DIY shoe assembly techniques.

-          Built a infrared-heated thermofolding machine sized to a single shoe in order to prototype foot plates at different thicknesses, materials, and foot bed profiles.

-          Built a 2-resonator analytical model for the transducer and foot bed suspension, informing the requirements of the shoe integration. Empirically matched the analytical model to buck-based testing, exploring sensitivity to different boundary conditions.

-          I identified a set of requirements for the transducer/shoe integration needed to ensure reliable and efficient resonance.

-          Extensively researched the tactile sensing modality and used existing analytical frameworks to translate product goals into quantitatively bounded requirements and specifications.

-          Worked with the CEO/product manager to categorize and bound the range of music targeted and then created a discrete number of reference tracks for an evaluation plan.

-          Built an instrumented haptic platform that could be reconfigured with different shoe suspension integrations, that measured vibration across reference footbed locations and could modulate signal processing parameters external to the embedded FW.  After an initial single-use experimentation phase where we mapped/explored relationships between different variables such as stance, power, filtering, construction, vibration amplitude at different locations, and signal content, we then went through a series of “listening/feeling” tests with different stakeholders in order to contrast the subjective evaluation to measurable variables, decouple signal content and system performance in the data, and identify perception thresholds.  A full mastery and understanding of all the variables were elusive, but the effort we did put in went a long way in helping to guide follow on design efforts and decisions.

-          Engaged a DSP/FW consulting firm to build a programmable signal processing toolset and protocol. Then worked with internal and external product stakeholders to adjust and tune the processing parameters through a hybrid subjective and objective process while ensuring we stayed within acceptable operating windows for power and buzz and rub.

-          Developed a process to measure frequency response and other relevant response parameters (white/pink noise & rub/buzz) when integrated within a shoe and in other WIP forms using a Keyence point laser and audio precision testers. We replicated this on a Listen tester and deployed multiple copies across the supply chain

-          The baseline design had multiple PCBs, batteries, and I/O’s integrated throughout the shoe with complex and susceptible wires and had been developed for a single shoe size. We consolidated the electronics into a single PCBA stacked with a cell after a trade study exploring dependencies between battery capacity, XY extents, and total module height.

-          I designed a unified enclosure with glass filled nylon and an aluminum cap that had strain relieved and sealed interconnects to a power button and charging port module that was integrated with the shoe upper as well as the inertial transducer and two antenna subassemblies integrated on polyethylene carriers.

-          Through the core functionality testing described above, we identified that an injection molded insole sized to the extents of the footbed was necessary to distribute the vibration and provide a consistent ground reference. Although this approach required a separate part for each shoe size and side, it allowed for the suspension stiffness to be tuned for each size as well and we controlled intentional diaphragm features in each foot plate to that end.

-          I developed and validated design specifications using a combination of analytical modelling and measuring force/ pressures in mockups and used structural FEA to optimize and proportion the structure. I also built a pneumatic test fixture to load a proxy foot and measure the static deflections across the outsole to confirm the compliance was in an expected range.

-          Because of the maturity of the transducer magnetics, suspension design, and supply chain we wanted to avoid as many changes as possible. We ended up modifying the attachment feature to make it fundamentally more tolerant and stiff than the baseline. We also identified that suspension failures previously attributed to uncontrolled resonances were in fact tied to incidental forces when kicking with or dropping the shoe. To mitigate this, we developed an internal guide/stop that was out of contact with the transducer through the operational window but engaged and stopped the field assembly before the suspension would break.

-          I engaged a pogo pin cable vendor to develop a 5-pin charging & data interface with a custom connector interface that could be elegantly integrated with the shoe upper. We mated the charge port with a backlit button and ensured the submodule could be integrated across different heel-cup sizes.

-          The baseline design used a simple wire for each wireless connection and was experiencing dropouts across most use-cases.

-          First worked a high value RF advisor to review the existing IC selection and PCBA design to create a starting link budget and set of initial requirements for passive antenna performance.

-          Built a system to evaluate the performance of reference mobile devices and headphones for the purpose of the link budget

-          With the advisor, I then did a trade study that modulated internal electronics/battery configuration and packing, as well as critical shoe-integration parameters (bond line and midsole proportioning) to converge on a starting dipole antenna integration considering ground separation, path loss, etc.

-          Engaged an RF design consulting firm to simulate the integration with and without a reference foot/environment and evaluate S11 values with a standardized receiver as well as isolation between antennas. And did some parameter sweeps/sensitivity analyses.

-          Built prototypes and manually tuned the matching circuit before testing the passive antenna S11 values with a spectrum analyzer in an anechoic chamber.

-          Booked the design for a system build and developed an RF test box that we configured for our modules and integrated shoe. Then took on the line measurements at the antenna, submodule, module, and shoe level. We the developed a series of functional tests and specification limits for use during follow-on builds and associated rel testing

-          Performance was improved significantly but dropouts were still a complaint that came up during beta testing after the build, so I then worked with FW engineering to build in the ability to track RSSI and include a button within the app to report a perceived dropout. Then I worked with QA to build a test protocol consisting of a controlled walk around several city blocks including several suspect features and with a specific content set. We were able to use the data to root cause most observed dropouts to uncontrollable environmental conditions and demonstrate parity with competitive devices. In a small number of cases, we were able to identify, root cause, and mitigate product related issues usually tied to assembly.

-          Another challenge we faced was that of latency in the audio chain which was a composite a multiple links in the chain, each with its own set of built in strategies to mitigate the perception of a dropout when the signal strengths drops. There were several initiatives to understand the different IC behaviors and

-          While we had an effective EMS partnership in place, there were limitations to the extent of support available, and there were not immediately viable partners and methods to manage shoe integration and final FATP. Additionally, we needed to develop a shoe design that met the product functional and comfort goals while being compelling to the target market and aligning with the brand design language.   

-          I immersed myself in the world of shoe design, development, and manufacturing in order to understand the workflows and constraints around midsole, outsole, and upper development traveling extensively to China. We also learned how to navigate the operational complexities of footwear such as last development and grading, quality control, and managing stock grades.

-          We engaged strategic footwear consultants and designers and studied a range of reference lasts and midsole formats and converged on a construction and reference last bottom that was synergistic with the needs of functional components. We then developed and styled an upper that was acceptable to various stakeholders.

-          We built a personal and business relationship with the outsole/midsole/insole mold supplier, unit sole molder and assembly house, as well two different factories to build the upper and do final assembly and packaging. We began production with a footwear factory that was well matched to our product and initial volumes with enough technical competency and motivation to adapt the process to the electronics integration including modifying standing rework and WIP test protocols.

-          We were able to contain the output of the EMS factory to three different SKUs (to support 10 different shoe sizes) and designed a (potentially) re-usable molded carrier on which the different functional elements were mounted. And developed a series of outgoing functional tests that ensured system and component performance.

-          Because the shoe factory did not have experience managing and handling complex electronics, we rented out an assembly line in a packaging factory close to the shoe factory. We would receive modules from the EMS in this location and assembly them to footplates according to the size demand, and then courier daily production volume to and from the shoe factory. I worked with FW engineering to source used equipment and set up a 12 station FATP line that could be configured to manage module IQC testing, footplate assembly and testing before integration and then shoe functional testing, final OQC, & pack out after integration. With enough volume we planned to develop secure storage at the shoe factory and then ultimately transfer the FATP stations.

-          Because production was not consolidated under a single CM we needed to navigate the operational implication of VAT and the intricacies of exporting effectively. Initially we partnered with a packaging company within an export processing zone to ship our products to the states.

-          In the US we worked with a 3PL to store our stock and fulfill orders direct to consumers and through channel partners. We built the capability for the 3PL to flash final firmware and ensure charge state efficiently.

Supply Chain

-          There is not a single product safety standard that was applicable to the product, and we were not prepared to pay the cost for UL to develop one. I identified that an active UL listing was not a barrier for our channels partners or securing product insurance, but built an internal product safety evaluation strategy based on that for IT equipment, UL62381 and augmented it to reflect the particulars of our use case, particularly in relation to battery safety concerns.

-          I worked with EE to carry out and submit FCC radiated emissions as well as BT SiG testing. Because the design did not use a prequalified module, the final testing was extensive and resource intensive. In order to derisk the effort, we strategically utilized pre-scans and internal protocol evaluations to ensure compliance and passing results.

-          Despite our best efforts, it was clear that the implications of manufacturing and distributing a shoe-integrated product were more burdensome and extensive than initially considered during early product definition. While minimized, the added weight of the components and resulting comfort implications kept full-day wear from being realistic for many customers, the Bluetooth connectivity limited operability to a mobile device, and the electronics cost drove a price point perceived as high for footwear. Additionally, it was difficult to identify a single shoe design that was widely appealing as lifestyle footwear.

-          In response to these insights, we conceived a complementary product targeted to in-home use with slides or similar house shoes. This product would be a replaceable insole allowing the functional electronics to be removed and swapped between shoes which greatly simplified the supply chain. Additionally, this product would function with a mobile app, work with a variety of inputs, be user adjustable, and allow the provision of a wired connection for time critical applications like gaming, watching movies, and live musicians. With the opportunity to design a transducer and suspension that maximizes sensation/power instead of prioritizing power at all costs, a longer battery life and smaller components could be used.

-          I developed a reference design and substantiating rationale and prototyped the slide integration using existing hardware and then nurtured an engagement with IAC in pursuit of an ODM development cycle. We got through scoping and quoting the engagement, but the company did not have resources at the time to undertake the project and focused resources on operations for the first product.

next gen 1

next gen 2

next gen 3

next gen 4

next gen 5