I’ve worked on projects at various stages of maturity, which call for different approaches depending on product-level priorities. These stages blend between one another, are often iterative, and the projects I’ve been exposed to have defined their phase segmentations differently, but here’s a generalised overview:
Conceptualisation & Definition
Proof-of-concept Prototype
Early-stage Prototypes
Late-stage Prototypes
Start-of- / Post-production
- Searching for product market fit
- Primary and secondary research to scope end-customer needs
- Ultra-low fidelity prototyping to demonstrate design intent
- MVP: defined product category and value proposition(s)
- Gathering feedback from early adopters
- Quick iterations with rapid prototyping and off-the-shelf components
- Early low-volume testing (engineering validation tests)
- Soft-tooled parts
- Top-down analysis: requirements listing and system decomposition
- Detailed function and manufacturing de-risking (design and production validation tests)
- First-off hard-tooled parts
- Pre-production sample runs
- Bottom-up tests: verification and integration
- Change management for mid- / post-manufacturing updates
- Detailed optimisation for cost-down and reliability factors
Presso ↗ develops rapid and distributed dry-cleaning automation robotics for the hospitality and real estate industries. The kiosk offers an alternative 24/7 garment care self-service experience for customers, whilst reducing energy consumption and labour costs for businesses that install the appliance within their facilities.
Stage
Product Definition, Proof-of-concept Prototype to Early-stage Prototypes
Responsibilities
Mechatronics Rapid Prototyping // Physical UX Design // Digital Product Design
0
1
I joined the two co-founders of Presso during my university industrial placement, whilst the project was being incubated at HAX, an accelerator program for early-stage hardware startups. As the first hire, I helped define the mechanisms and sub-system designs which allowed the machine to perform its function, and facilitated the transition from an R&D project towards a consumer-facing product.
Within fifteen months, the kiosk has undergone four major iterations (top to bottom, left to right):
a test rig to rapidly prove out concepts; a proof-of-concept unit that demonstrated functionality with user-facing features; 2x looks-like, works-like units for initial pilot trials; 10x trial units with improved robustness and reliability
The kiosk’s main functionality is to hold a garment’s fabric in a taut position, and allow pressurised steam mixed with disinfectants and scents to re-align the fibres within the fabric into a ‘ironed’ condition.
I designed sub-systems with mechatronic, hydraulic and pneumatic aspects, trying ideas with quick prototyping, resulting in refined concepts using a mix of 3D-printed and off-the-shelf components.
Sleeve pipes - responsible for gripping a shirt’s cuff, providing a moving steam outlet, and allowing for rotation as the sleeve was being pulled.
Trying to utilise conventional robotic end-effectors led to a realisation that they did not require a balanced centre of mass for free rotation along one axis. By offloading the actuation mechanism to the fixed base, and using a symmetrical, bearing-mounted and spring-loaded design, the gripper was able to perform its function.
Hanger with inflatable pods - responsible for separating a garment’s interior volume into three ‘pockets’ (body and two sleeves).
Heated blow dryers - the drying system of the machine.The kiosk has an existing heat source with its pressurised steam chamber. By swapping out the nichrome wire heating elements within hair dryers for heat exchangers, this design was able to reduce the machine’s energy consumption.
As users were responsible for securing the garment within the kiosk, it was important to provide a user experience that was intuitive to understand and pleasant. I prototyped the physical touchpoints and touchscreen UI which provided affordances towards user interactions.
Initial user-journey mapping to layout machine activities against user actions and information load.
Through low-fidelity ‘Wizard-of-Oz’ prototyping with a mock-up kiosk and sketches of UI instructions, testing and quick iterations were conducted with first-time users.
Chosen instruction methods were developed into higher-fidelity renders.
During COVID, customer trials were pivoting towards film & TV studios who wanted quick-turnaround garment-cleaning for their costume departments. As such, a modified user experience and touchscreen interface had to be designed towards costume assistants, a "power user" persona who constantly used the machine. It was also a valuable opportunity to rapidly collect feedback on the machine’s performance via a quick survey with net promotor scoring.
A dashboard had to be developed for monitoring and analysing machines used for customer trials, happening remotely across different cities. I designed the dashboard UI for reading and displaying back-end information from trial units which were outputting sensor data and camera feeds.
These mockups were used as input by the software team to create a JS-based web-app.