Zoos
This note tracks the PhD project on animal-led housing across the landscape of captivity (Kathy Chapman, Deep Design Lab, from 2027). The topic is larger than the application, so this note also collects cases, methods and open questions for later lab work.
Files: Deep Design Lab - Documents\Research\Zoos\Argument (argument outline, Chicago author-date, Typst output), ...\Zoos\PhD (proposal drafts, CV, requirements) and ...\Research\Owls\Projects\2025 Zoos Victoria - Leadbeater's possum hollows (YES) (the hollow project that supplies one confirmed case).
Collaborators
- Mia Cobb, Animal Welfare Science Centre, University of Melbourne: prospective second supervisor; co-convened the Kennel Hackathon; working-dog welfare, the agency domain, welfare through competence.
- Sarah Webber, School of Computing and Information Systems, University of Melbourne: lab collaborator in animal-computer interaction; co-design with orangutans; prospective advisory or co-supervisory role.
- Zoos Victoria (Dan Harley, Darcy Watchorn, Arabella Eyre): the lab's contract for 3D-printed Leadbeater's possum hollows. The design work is the lab's (Roudavski and Parker) and is complete: Phase I (2025 to 2026, $15k) produced six small-scale and two full-scale prototypes against the keepers' brief (fixed 20 L dimensions from the old recycled-plastic boxes, 5 cm entrance, hinged lid at 40 to 45 degrees, pins or bracket). Phase II (proposed 2026) is a production series of 40 with optional validation by captive possums at Healesville; Phase III (2027) is field validation; Phase IV is a build-test-learn cycle. Zoos Victoria plans about 4,000 hollows over a decade. This design work is not part of the PhD. What the applicant could do, to be confirmed with the partner: iterate on the designs, try new materials, integrate possum observation with experimental variants in captivity, learn from and integrate field monitoring.
- Deep Design Lab (Dan Parker, Alexander Holland): workflow, facilities, precedents.
Existing collaborations count with assessors, so the proposal lists them and the portfolio contains the evidence (letters, project pages).
Related Notes
- Settings: Zoo, Captivity, Kennels, Domestication, Companion Animals
- System and object: Housing, Habitat Structure, Artificial Habitat Structur
- Agents: Agency, Agentic Stair, Biological Individuality, Capabilities
- Participation: Ladder, Leadership, Participation, Interspecies Design
- Methods: How to Write and Use an Argument Plan, More than Human Design, Animal Computer Interaction
- Justice: Justice, Thriving, Health
Argument in Three Parts
- Landscape of captivity. Housing must be understood across the whole landscape, from companion households, kennels and laboratories through zoo exhibits, sanctuaries and soft-release pens to fenced havens, islands, liminal urban settings and wild animals with artificial structures. The Kennel Hackathon (1 September 2026) is one sample on this spectrum.
- Agents rather than species. The unit of design is a declared decision: individual, pair or colony, population across institutions, species with its microbes, institution of animals and carers, landscape. The five declarations from Agentic Stair (boundary, variable, horizon, perspective, beneficiaries and costs) keep cases comparable.
- Animal-led co-design. Animals lead through use, choice and refusal of adjustable housing in iterative cycles: propose, build adjustable, animals assess, measure, revise. The object is a housing system, rather than an enclosure, that supports thriving and justice over time.
Methods run from breadth to depth: evidence review and expert consultation across 8 to 12 cases, personas and sketch designs across many cases, prototypes with the animals in two or three settings, full-scale living labs in one or two, evaluation against the ladder and capabilities, scenarios at one, five and fifty years.
Focal Cases
| Agent | Settings on the landscape | Partners to approach | What the animals could decide |
|---|---|---|---|
| Eastern barred bandicoot (Perameles gunnii) | Breeding facility, fenced havens (Mt Rothwell, Woodlands), fox-free islands (Phillip, French, Churchill), farmland trial sites with guardian dogs (current status to confirm) | Zoos Victoria, Mt Rothwell, Phillip Island Nature Parks | Mates (shown to help), refuges, routes, nest sites, exposure to predator cues |
| Brush-tailed rock-wallaby (Petrogale penicillata) | Breeding program, fenced reserves, Grampians and Snowy River releases | Zoos Victoria, DEECA | Shelter sites, social grouping, release timing |
| Leadbeater's possum (Gymnobelideus leadbeateri), in discussion | Captive possums at Healesville Sanctuary for trials of material and detail variants of the lab's existing hollow; installed hollows in Central Highlands forest (field monitoring from 2027) | Zoos Victoria (Harley, Watchorn, Eyre), DEECA | Selection among variants, entrance use, colony composition, thermal conditions; the brief's fixed dimensions are a human-led baseline the possums' responses could revise |
| Dog (Canis familiaris), confirmed case | Working-dog kennels (Animal Welfare Science Centre partners), shelters, households, free-ranging | Animal Welfare Science Centre (Mia Cobb), Kennel Hackathon participants | Doors and partitions, companions, contact, rest, routines |
Further cases are to be established with partners: a shelter, a laboratory colony, a zoo exhibit, a sanctuary, a soft-release pen, a fenced haven, a fox-free island. Candidates from the literature: wombats under treatment for mange (Foth et al. 2026), kittiwake hotels (Wilson et al. 2026), possums in suburban roofs (Power 2009).
Open Tasks
- Add the eight references cited as placeholders to Zotero and re-export the bib files (done 6 October 2026).
- Optionally add the further suggestions at the end of the argument outline (foundational zoo design, agency in welfare science, haven debate, bandicoot and rock-wallaby recovery literature).
- Convert the hand-typed references in Captivity and Zoo to Chicago 17th footnotes rendered from Zotero.
- Draw Figure 1, the landscape of captivity, as the front-page figure in place of species photographs.
- Confirm with Zoos Victoria whether and how the applicant can work downstream of the lab's hollow design (iteration, materials, captive trials, field monitoring).
- Confirm partner facts before they enter the proposal: Zoos Victoria's guardian-dog program for bandicoots, current haven and island populations, Leadbeater's possum captive population status, Sarah Webber's role.
- Guardian dogs, checked on 8 October 2026. A 2017 review confirms that Zoos Victoria began training Maremmas in 2015 and planned trials with bandicoots, as cited in Captivity. A 2021 project factsheet reports that researchers released 20 bandicoots at each of two farms, Mooramong in November 2020 and Dunkeld in June 2021. The factsheet attributes two deaths to foxes and notes that other bandicoots went missing, while several surviving females bred. No journal article reports the outcome yet, so ask Zoos Victoria whether bandicoots still live at either farm.
- Revise the proposal docx against the outline and its word budgets (2,500 words); assemble the portfolio to the structure in the outline; complete the CV (100-word summary) and the supervision form (co-supervisors under additional comments).
- Decide which cases get prototypes and which get living labs once partners respond.
Open Questions for the Lab
- Which welfare and agency indicators apply across settings from zoo to haven, so that one evaluation can compare them?
- When does added choice legitimise confinement without changing its terms, and how would a design show the difference?
- Can exit (release, rehoming, transfer) be designed as an option within a housing system rather than as its end?
- How should the method weigh the preferences of one resident against the welfare of its group or of other species, including the predators excluded by a fence?
Backlinks