MichaelZhang
Scope Research

Defining STRUTT ev¹

How might we define a mobility product around what people actually need, when the company started from technology rather than a customer?

The company began with technology rather than a customer. We had a capable sensing and algorithm stack and a long list of things it could do, but no reliable picture of who would use a powered mobility device, what they actually cared about, or why. Product definition was being argued from intuition, and the arguments could not be settled because nobody had the evidence.

I led the user research, designing the study, running the interviews and home visits, and synthesising the findings into the definition work. I worked directly with the design director and reported to the CEO. Later in the same phase I built a Grasshopper tool that let the team see real sensor coverage instead of estimating it from overlapping geometry.

Research goal

01  Who our users are
02  How they actually live
03  Whether our assumptions hold

Research process

We recruited online for remote interviews, which let us reach people across different ages and countries. We also recruited locally, so we had people we could meet in person, observe, and test prototypes with.

01 Who our users areWe recruited across the full range of mobility need, from people who occasionally use a cane to full time electric wheelchair users, and across a range of ages and countries. We kept the sample broad to begin with and narrowed it later, once we understood the groups well enough to know which ones we were actually for.
02 How they actually liveWe built a set of methods for walking people through their own days in as much detail as they could give us, covering grocery shopping, work, travel, moving around inside and outside the house, and getting about in cities and in smaller towns. The more we knew about how someone actually spent a day, the easier it was to understand why they made the choices they did.
03 Whether our assumptions holdWe asked about bad experiences, about the inconveniences of ordinary days, and about what people had already done in response. We tried to stay on facts rather than opinions here, asking what had actually happened rather than what people thought about it. Only afterwards did we look at whether our technology could have helped in that particular case.
Me interviewing users online

Research outcomes

We completed more than 10 stages of online interviews with more than a hundred participants, and dozens of site visits in person.

01 Who our users are We narrowed to three target users.
  1. The experienced user. Over sixty, ambulatory but able to walk less than a hundred metres, with moderate mobility loss and prior experience of a powered mobility device.
  2. The first time buyer. Over fifty, with mild to moderate mobility loss. Able to walk, but slowed by balance problems or declining stamina. Typically using a walker or rollator, and considering a powered device for the first time.
  3. The family buyer. Aged roughly forty to fifty, buying for a family member so that person can rely less on others and be independent again.
We also decided who we were not designing for. Buyers looking for recreation, and people whose needs a complex rehabilitation chair.
02 How they actually liveMost were retired. Their weeks were built around grocery shopping, time with family, walking around the neighbourhood, going to church, and travel, including cruises and theme parks. Many had spells of being housebound because of their mobility, but they did not want that. What came up again and again was wanting to go out independently and enjoy life the way they used to.
03 Whether our assumptions hold People consistently described scraping or knocking into things in narrow spaces, running into others in crowded ones, and getting stuck or losing control on uneven ground. Some had responded by avoiding those places altogether, which was not what they wanted. That is the situation our sensing and navigation work is meant to help with. A second theme came up alongside it. Many people felt self conscious on a device that looked medical, and said it made them feel disabled. Some were already adding stickers or coloured LEDs to make the device feel more like theirs. What they wanted was not decoration, but a device that did not look medical in the first place and looked capable instead.
Me interviewing users in their home

Making sensor coverage visible with
Rhino & Grasshopper

While the user and product definition work was running, the R&D team was debating where the LiDAR sensors should sit and how many were needed. It was not only a question of technical blind spots. Coverage also depends on the size of the person in the chair, how they sit, where they put their things, and how they transfer in and out. At the time the team was working from hand drawn illustrations and overlapping 3D models, because there was no accurate or quick way to check what a given arrangement would actually see.

I wrote a Grasshopper program for LiDAR field of view inspection. The algorithm and design team could move each sensor, change its orientation, adjust the beam angle, and drop different human models into the chair. The program calculated the blocked field of view in real time and sliced it by height, so coverage could be read separately at ground level, at desk level and at head level.

It helped the team agree on the sensor placement, and it gave design and engineering something concrete to work against. A proposal could be checked against a real body in the chair in seconds, so form decisions and sensing performance could be discussed together rather than separately.

The field of view of 2 LIDARs in the front
The field of view at ground level