5

System

Images borrowed from Silky Szeto

Packaging + Accessories

Images borrowed from Silky Szeto

Overview

This was the most expansive design challenge I've undertaken. To call this hearing device ambitious would be an understatement. Aside from internal hardware electrical components, the only non-custom item is the wall plug. The R&D team had to develop several complex systems from scratch - a completely new magnetic wireless miniature charging platform, a new hearing device architecture built around higher performance hardware, two new tip styles, a completely new assembly and disassembly method ... the list goes on and on, so see below for more details. Balancing all the design constraints across these systems was quite tricky, so consequently, I'd consider this project the pinnacle of my technical achievements thus far.

This project also forced growth outside of design activities. Being part of the leadership team throughout the project, I learned to adapt quickly as the project, team, and company evolved. Aside from pure mechanical design, I actively participated in defining scope, vetting requirements, iterating industrial design, rendering to support user guides and marketing materials, packaging and graphic deisgn, selecting and activating a new supply chain, hiring and managing the supporting team, and divising methods to increase team agility.

All said and done, the result is pretty remarkable. It's hard to convey the actual scale of how small this device actually is in real life - the entire hearing aid is about the size of my fingernail. I'm also really proud of the details in this design - which in my opinion, combine to create the overall impression of a high quality and thoughtfully considered product.

Design Challenges

Miniaturization. Particularly in the early prototype phases, the hearing device architecture required extensive optimization to reducing size, and ultimately improve comfort for smaller ears. This centered around early hardware selection - namely selecting a new DSP / microphone / receiver / accelerometer combination and power budget estimate, in order to create a customized miniature battery. Once this hardware was locked, an iterative hardware Tetris process began, followed by surfacing and enclosure design. The plastic design was quite agressive - the nominal thickness of the enclosure was 0.3mm, diving as low as 0.1mm in critical areas - requiring extensive collaboration with our tooling design partners.

Assembly and disassembly. During initial development, Eargo was seeing about 35-40% return rates for existing products on the market. This figure is about average for the hearing aid industry, where physical and lifestyle fit are harder to predict - but returns cannot be resold due to biohazard contamination risk, and disassembly was impossible due to the status quo glue-based assembly method. Unlocking a way to disassemble and harvest the expensive internal hardware components became a top priority requirement to increase profitability. Laser welding was the key. More specifically, laser unwelding an enclosure with the internals suspended within was the solution. And as far as I know, this was also an industry first.

Sustainability. Hot out of grad school at Berkeley, I was fired up to apply my new knowledge of sustainable design techniques. Aside from the return harvesting discussed above, this manifested in a number of other creative ways - the hearing device enclosure is made from bio-based nylon and compostable BioPBS, all the packaging is recyclable, and the entire product architecture was devised to be as minimal as possible to cut down on both part and tooling footprint. Accessory packaging was designed to utilize shared components where possible to communize tooling. To reduce warranty claims, the entire system was designed to withstand three years of abuse. To mitigate wax failures, we even created a custom cleaning tool and a clever replacement tip tray pack to make regular maintenance easier.

Wireless charging. All previous Eargo products were charged using simple electrical contacts on the hearing device, which were sensitive to wax contamination and sometimes finicky for the target demographic to correctly dock in the charger. To enable an easier and more robust charging experience, this interface was completely gutted and replaced with a miniature wireless induction charging system with magnetic device retention. This system took a few tries to get right, and efficiency was the ultimate challenge. Maximizing the coupling between transmitter and receiver coils became the priority - pushing us to devise creative ways to reduce molded plastic thicknesses to as low as 0.1mm, experiment with the minimum thicknesses of PSAs able to withstand water ingress, devised blind assembly methods to close coil gaps, and conduct extensive magnetic FEA simulations to reduce field interference.

Tips. The tips were required to have positive haptic feedback, so customers could be sure they were safely assembled to the hearing device. We also needed two versions - an open style with more airflow, and a closed style with more acoustic occlusion - in three sizes. After conceiving the initial biomimicry inspired industrial design concept for each tip, it took a huge experimental study to nail down the final configurations. We scanned and statistically analyzed ear forms. We extensively used both FEA simulation and physical testing with the clinical team to iterate each aspect down to the smallest details - tweaking fractions of a newton in snap force, dialing the petal surfacing to reduce silicone tear initiation points, and so on. Even after final design lock, we had to troubleshoot a micro-cracking issue at a weak melt front on the plastic insert, resolved by a simple gate size increase and adjusting molding process. The result was all-day comfort, a safe and satisfying snap, with adjustable occlusion and fit - all in a captivating form factor.

Feedback. If the microphone picked up either acoustic or mechanical vibration from the receiver inside the hearing device, it would throw itself into a fatal feedback loop. Prior architectures solved this using encapsulation glue, so a new solution was required to enable disassembly harvesting. Ultimately the team devised an internal vibration dampener system - kind of like a miniature suspension surrounding the receiver - combined with a stack of robust PSAs seals surrounding each enclosure port.

Water ingress. Previous Eargo products also suffered from water related returns - people would forget they were wearing the hearing devices, and hop in the shower or go for a swim. To combat these failures, we extensively studied all hydrophobic solutions - including membranes, hydrophobic coatings, and hydrophobic meshes. Ultimately the best solution was a super hydrophobic acoustic mesh combined with specific adhesives, with a specific weave pattern that suffered less from wax buildup.

Wax ingress. After purchasing, the top return issue was earwax buildup. Both the output and input ports required improved wax protection, so we devised new semi-passive solutions for both. At the front, a simple torturous path prior to a hydrophobic mesh provided adequate protection, and enabled customers to easily remove and clean the primary buildup areas. Late clinical testing found that our microscopic wax guard (an injection molded of similar scale to a speck of sand) did such a good job protecting ingress that it would clog quickly - so it was simply removed to provide a more open interface for regular maintenance. At the back, the mic port was protected by a separate disposable, compostable consumable accessory Mic Cap - which even included microscopic rotating wiper blades to scrape up wax before being disposed.

Cost and scope. The charger initially included buttons and speakers to ultrasonically control the hearing aids, and a premium external paint coating to completely hide the indicator lights and provide superior scratch resistance. The buttons, in their floating unsealed industrial design, rendered the charger unable to meet water resistance requirements. Moreover, given our limited volumes, our BOM cost target was unachievable and paint was an extravagant cost driver. These forced two massive redesigns to the charger architecture to remove buttons in late validation builds. The takeaway lesson here is to focus early on minimum viable product and remove all unnecessary complexity at the initial conceptual phase.

Pull tab optimization. The pull tab on the initial hearing device concept was attached to the tip, intended to avoid any possibility of tips falling off in the ear. Lack of clinical testing on early prototypes missed a huge safety risk - in larger ears, the hearing device could hook itself in the first bend of the ear canal, and force applied to the tip pull tab was unable to safely remove it. This forced another late redesign to the entire hearing device and charger cradle, to move the pull tab to the rear of the hearing device. Towards the end of the project, additional delayed testing results required further optimization of the pull tab to improve grip feature tactility and overall length. The takeaway lesson is proper testing is critical - major risks must be validated as early as possible in the most realistic test scenario available.

Emissions. The entire system was not fully combined until later validation builds, and EMC testing failed late in the project. Luckily, the team was able to resolve the issue with a simple addition of an EMI choke component to the USB input.

Controls. Previous Eargo products utilized an acoustic tap to change amplification programs while wearing the hearing devices - in other words, a user had to gently slap the side of their head to increase volume. A simple addition of an accelerometer to the hearing device changed this slap to a tap of the tragus, significantly improving the user experience and control reliability. More impressively, the team devised a novel method to control the hearing devices ultrasonically via a phone app. This ultrasonic system was also leveraged in the charger to create a communication backchannel during charging - the hearing devices could talk to a mic in the cradle to confirm updates / halt charging / etc.