Prototyping the cat-centred device

This stage of the research had the aim of designing a prototype derived from the workshops requirements to show that the wearer-centred framework (WCF) could be operationalised and to demonstrate that, by means of it, a design centred on wearers could be achieved. We had two possibilities: to either choose one of the three requirement sets emerged from workshops as a whole or select requirements from all those established by the three teams. We opted for the second option based on the fact that the focus groups were design trials limited by time and participants’ competences, therefore each set taken individually was not complete enough (arguably, in a real case scenario, where teams would include multiple expertise, this effect would presumably be much reduced). For example, the set of requirements established by the computer scientists (Table 6.1) missed accounting for features such as an aerial, or aspects such as the emission of light spectra or odour. Therefore, to systematically choose the most appropriate requirements of the three sets for any of the feature or quality listed in Table 6.1, we proceeded as follows. Firstly, we selected all the workshop requirements that were in common across the three teams (from List A in section 6.1.4). Then, where differences applied (i.e. List B and C), we chose one or another team’s workshop requirement, depending on the extent to which designers reflected about wearability; that is, if and how teams considered wearability implications and expressed aspects accounted for by the framework. For example, the computer scientists specified that the device should be a unique piece of elasticated material in order to avoid the collar scratching or chaffing the cats’ skin, while the biologists and cat carers opted for standard collars to be easily fastened through closing mechanisms. In choosing which kind of collar could achieve a better wearability, we selected the unique-piece elastic band since the computer scientists were concerned about the discomfort that any fastening method could exert anytime on the wearers, while the biologists and cat carers were more worried about the momentary problem of putting on the collar. In other words, the former had a more accentuate wearer-centred perspective than the latter. Looking back at the study-1 requirements, the computer scientists’ requirement we chose is consistent with the empirical outcome that fastening methods such as Velcro or buckles should be avoided (see empirical requirement of feature e- in Table 6.3), which further validates our choice.

However, when the differing requirements had equivalent wearability orientation, we opted for any same requirement established across two teams, since we deemed that selecting the same outcome reached independently by two teams would constitute the most objective criterion, due to the inapplicability of the other two criteria. Finally, the workshop requirements identified by only one team (in List B) were directly selected. Schematically, for each feature or quality listed in Table 6.1, the requirements selection approach followed 4 rules:

1- Select workshop requirements identified across the three teams (these are summarised in List A);

2- If workshop requirements differ across the three teams or are similar across only two teams (these are summarised in List B and C), select the one that expresses greater consideration for wearability issues;

3- If workshop requirements are similarly oriented towards wearability, select the one identified across two teams;

4- If a workshop requirement is not identified by more than one team, select the one identified by one team.

Limitations of the approach

This systematic approach to the choice of the requirements was employed to develop a prototype derived as much as possible from designers’ insights. However, this led to some prospective flaws in the prototype. In one circumstance, a requirement was selected to adhere to the above protocol (rule 3) even though we thought that another requirement would have been more appropriate. In this example, the biologists were concerned that a wireless charging system could produce heat that could potentially harm the wearer, so the team opted for the removal of the batteries during recharging. Although this is a sensible idea, we stuck with what the other two teams had established in relation to this design feature, that is recharging the batteries wirelessly, since both biologists’ requirement and the other two teams’ requirement were established with wearability issues in mind: avoiding the device overheating for the biologists, and reducing the burden on the wearer for the computer scientists and for the cat carers.

7.1.2 The prototype

Following the approach described in the previous section, we sketched the cat-centred prototype illustrated in Figure 7.1. However, when it came to make the actual physical device (Figure 7.2), some of the features were modified in accordance to the resources available to us and the feasibility of implementing what the workshop designers had proposed. This resulted in a prototype that partially differed from the one sketched in Figure 7.1, as discussed below.

Sketched prototype

As per established by all three teams (see List A in section 6.1.4 and Table 6.2 for reference) the device (Figure 7.1) had to be a narrow built-in collar; not protruding inwardly and minimally protruding outwardly; weighting equally or less than the lightest device on the market; colour blended with individuals’ fur; easy to pull out; retrievable; with soft texture; wrapped in a thin waterproof coat; with thin and narrow components distributed along the collar and aligned end-to-end, connected to each other through some conductive material. These requirements were according to rule 1 of the prototyping approach described in section 7.1.1.

Figure 7.1: The sketched prototype and its components.

Regarding the number and sizes of the technological components, we sketched them relying on the same assets in the PawTrax® Halo tracker tested in study 1, which consisted of two batteries, an integrated GPS/GSM unit, an antenna, and a charging element. As mentioned in chapter 5, section 5.2 (p. 76), at the time of this research, this was the lightest GPS available on the market (weighing 21.7 grams).

Workshop requirements among Lists B and C (section 6.1.4 and Table 6.2 for reference) were selected as follows:

As proposed by the computer scientists’ team, we opted for a unique piece of elasticated collar that gets pulled on and off the cat’s head without opening and closing the collar extremities. This solution was chosen to avoid fastening mechanisms that might somehow irritate the cat’s skin and it is consistent with empirical findings that buckles and Velcro straps might catch fur and prickle hair follicles. This feature also affords safety since a low- tension elastic textile pulls easily off the neck if stretched (this was assumed by the team

and tested with the principal investigator’s cat). This design choice followed from rule 2 of the prototyping approach (section 7.1.1).

Although the use of soft and flexible material was established by the three teams, there was no agreement on a specific material. Following the concern from biologists that devices should be odourless (also equivalent to an empirical requirement of feature f- in Table 6.3), we dismissed the use of silicone or rubbery material, which might have strong odours. Instead, we stuck with fabric as proposed by both computer scientists and biologists. Both rules 2 and 3 were followed.

We did not have agreement across teams about how to recharge the batteries with one team (the biologists) proposing that they should be detached when out of power and the other two teams (the computer scientists and cat carers) positing that the cells should be charged wirelessly. Since both detachability and ‘wirelessness’ were similarly wearer-centred (i.e. to decrease the device’s intrusiveness and avoid its impact from overheating), the sketched device was designed to have a radial wireless charger according to what was identified across two teams (i.e. the computer scientists and the cat carers). This followed from rule 3. Any sound actuator was avoided following the requirement of computer scientists and biologists that acoustic signals should be avoided (consistent with rule 3), and the requirement of the biologists that visual and osmic elements should also be avoided (consistent with rule 4).

The cat carers’ team was the only one to address the matter of the antenna, therefore we used their requirement that there should be a threadlike aerial along the collar. This followed from rule 4.

Actual prototype

The actual cat-centred prototype is illustrated in Figure 7.2. Its technological components are those of the dissembled PawTrax® Halo device used in study 1 (as mentioned earlier, these were chosen to keep the collar as light as possible). Specifically, such components were two lithium batteries (3.7 v, 160 mAh), a micro USB port, a switch, a customised GPS/GSM unit and an antenna (from left to right in Figure 7.2a). In the original product, the components were wired together and kept side by side inside two rigid boxes (Figure 7.3). For our prototype, we disconnected and re-wired them together to evenly distribute them and thus allow flexibility (Figure 7.2a). Then, we wrapped the electronics inside a thin waterproof coat (Figure 7.2b) and placed the wrap on a 9mm-width elasticated band, which we covered with a textile (Figure 7.2c). In this way, the elastic band was inserted into the fabric wrap which could slide along it. Finally, the two elastic band’s edges were sewed together to make a collar (Figure 7.2d) and the seam was slid under the textile cover in order to hide any discontinuity of the band’s inner line that might prickle the skin. Although the prototype was not a working device, using real components was deemed important to

simulate as much as possible features such as the weight and size of a real device that would matter from the perspective of the wearer. Figure 7.2e shows the prototype attached to a real-size stuffed cat toy (the same dummy cat used during the workshops).

Figure 7.2: The actual prototype – a) the components wired together to allow flexibility, b) the components wrapped inside a thin protecting layer, c) the wrap is covered with textile, d) the completed prototype, e) the prototype worn by a stuffed cat.

When designing the actual prototype, we tried to follow as much as possible the sketch in Figure 7.1. We implemented the concept of a built-in collar made of a soft and stretchy textile; adopted the solution of an unclasped collar in the shape of a hoop; kept the overall device as narrow as possible by choosing a narrow elasticated band; coated the electronics with a thin protecting film; distributed the components along the band as much as possible and minimised their inner protrusion. However, due to feasibility issues that emerged while making the collar, we had to trade-off the following features:

The idea of having the component spread at equal intervals all around the collar had to be modified due to the difficulty of crafting a complex stretchy design. For example, we could not find stretchy but resistant conductive material to connect the electronics such as coiled or elasticated wires, or elasticated conductive tape, or conductive ink resistant to pulling stress. Hence, we used normal wires to connect all the parts together, resulting in a narrower distribution of the components contained in a flexible but non-stretching section, connected to a ‘naked’ elastic band that provided the stretchy function.

We did not have the availability of a threadlike aerial. Thus, we used the rectangular one obtained from the PawTrax device disassembled.

The wireless charging transmitters available to us were too big and heavy to accord with the requirements of keeping weight and size of the device at a minimum. Thus, we opted for a standard mini-USB charging port.

The resulting device was then evaluated with cats in study 2, as reported in the next section.