Wildlife care / A closer look

What’s New in
Zoo Technology

The useful revolution is in what a zoo can know about an animal.

By Rob Urban · Research reviewed October 7, 2026

A heat map of an animal’s habitat looks persuasive. It has colors, contours and the apparent authority of a scientific instrument. Before it influences a habitat decision, someone needs to ask what those colors actually mean.

The heat map is where the story gets interesting

ZooMonitor’s own reporting documentation explains that the colors in its heat maps are relative to the data plotted. They do not represent fixed levels of space use. Its behavior reports also depend on the sampling method and whether the observer is recording an individual or a group. That detail belongs near the beginning of a technology discussion, because a polished visualization can obscure the choices that produced it.

Consider a hypothetical habitat review. One image suggests that an animal spends much of its time near a particular doorway. That could prompt questions about feeding, access to an indoor area, social relationships, shade or the timing of observations. The picture does not choose among those explanations. If observation periods mostly coincide with a particular husbandry routine, extending the sample may change the apparent pattern.

The practical achievement of structured observation is that the team can return to the question with a consistent method. A keeper’s concern becomes something colleagues can investigate together. A habitat modification can be followed by another period of observation. The zoo builds a record of how it reached a decision, including the occasions when its first explanation was incomplete.

Lincoln Park Zoo describes concrete uses of ZooMonitor: evaluating pygmy hippo behavior around enrichment changes, following chickens’ adjustment to a new habitat, and recording reproductive and dominance behavior in Japanese macaques. Trained staff, volunteers and interns observe more than 100 animals at the zoo. The app has been publicly available since 2016. This is established practice that helps explain the current direction of zoo technology, rather than a newly invented 2026 capability.

What adoption actually asks of the animal care team

A zoo introducing structured observation has to agree on a behavioral vocabulary before it can compare records. “Resting,” “inactive” and “not visible” need usable definitions. Observers need practice applying those definitions to the same situation. The team also needs a sampling schedule that answers its question without quietly substituting the times when volunteers happen to be available.

My suggested first project would be a question the care team already wants answered, such as how use of a habitat changes after a specific modification. Record the observation method, relevant husbandry events and visibility limitations. Review disagreements among observers before treating the resulting chart as a baseline. The deliverable should be an informed care discussion with traceable evidence. Counting observations alone would reward activity without showing whether the zoo learned anything.

Two different jobs for cameras

Seeing an animal.
Recognizing a species.

These are often collapsed into “AI for wildlife.” Their evidence requirements are different. An image classifier can help sort a field survey without being qualified to interpret an individual animal’s welfare.

The purchasing brief should name the actual job: find relevant footage, identify a species, distinguish individuals, recognize a defined behavior, or support a care assessment. Success at one does not establish success at the others.

The new camera problem is abundance

In its April 2026 account of conservation technology, San Diego Zoo Wildlife Alliance describes using AniML to help process trail-camera images. The team trains different machine-learning models for different regions. One Vietnam survey produced more than 300,000 images despite severe weather. The useful development is the ability to organize material that would otherwise create an enormous review burden.

That is a field-conservation example. It is not evidence that the same model can diagnose behavior in a zoo habitat. A zoo considering computer vision should ask a supplier to work with representative footage from the proposed installation. A demonstration filmed in an unobstructed enclosure at midday leaves unanswered what happens with reflections, vegetation, several animals crossing, low light or an animal outside the camera’s view.

A sensible evaluation keeps those difficult conditions visible. Review false identifications and missed events separately. Ask whether the system records uncertainty and retains the footage needed to investigate a result. If it identifies individuals, examine how the team corrects a mistaken identity before that mistake becomes part of an animal’s history.

Continuous video also changes the staff workload. Someone must decide which clips deserve review, how long material is retained and what happens when a camera moves. The economics depend partly on whether the system reduces searching or merely produces another queue. This is where modest tools can be valuable: a searchable archive of observations may solve the immediate problem without pretending to automate the care judgment.

Beyond the habitat

One animal.
A history that travels.

A camera records a moment. A zoo also needs to know which animal that moment belongs to, where it came from, what happened previously and which other records should follow it. Record continuity becomes especially important when care crosses departments or institutions.

Species360’s June 2026 ZIMS update moved rescue and collected-from-wild accessions into its newer interface and retired the corresponding legacy route. It is a specific workflow improvement in an established information system. The announcement supports a claim about the interface change, not a quantified improvement in rescue outcomes.

For a zoo evaluating a records change, the revealing demonstration is an awkward case: an animal arriving with incomplete history, an identification corrected later, or records transferred between teams. Watch where uncertainty is recorded and how corrections remain visible. A tidy example with every field completed will not reveal much about the work the registrar and care staff actually inherit.

The record now has another possible future. In August 2026, San Diego Zoo Wildlife Alliance and Species360 announced a partnership to connect biobanking information with ZIMS animal health, care and population records. The plan would make preserved biological material easier to locate in the context of the animal it came from. The source describes development of a shared system; it does not establish a completed global rollout.

This is a substantive horizon for zoos. A stored sample becomes more useful when its identity and accompanying information remain intelligible years later. For institutions considering participation, the immediate work is a conversation among veterinary, collections and conservation staff about identifiers, sample descriptions, permissions and the quality of existing records. Buying a visitor-facing AI product would do little to prepare that foundation.

The hospital is part of the technology story

A focus on software can make the physical side of zoo technology disappear. A June 2026 account of the Harter Veterinary Medical Center describes CT and ultrasound alongside patient handling, recovery spaces, environmental controls and camera monitoring. The article marks 25 years of the center’s work. It documents an integrated care setting, rather than announcing that CT suddenly became available to zoos.

The capital-planning lesson is in the relationships among those systems. A diagnostic capability occupies space, requires trained people and fits into a sequence of handling and recovery decisions. Its value cannot be evaluated as though it were a standalone device on an equipment list. For medical requirements, the institution’s veterinary team must lead the specification.

For leadership and fundraising, the communication opportunity is more grounded than “a high-tech zoo.” Explain the particular capacity being developed, why the care team needs it and what supporting infrastructure the funding includes. That gives a donor something more meaningful to understand than a screen or machine photographed at a ribbon cutting.

What should the visitor be allowed to see?

The public does not need unrestricted access to internal animal records. It does need a more legible account of the institution’s work. Technology can help a zoo show why a habitat includes retreat space, how observation informs enrichment, or how a conservation partnership extends beyond the animals visible during a visit.

This calls for deliberate translation. A short story about a question the team investigated may be more useful than a public dashboard of unexplained measurements. A dated update about a conservation project can acknowledge that research is continuing. A keeper-approved explanation of animal choice can help visitors understand why visibility is not the only measure of a good habitat.

There is also a straightforward digital service obligation. If a habitat is temporarily inaccessible, an encounter has changed or a route is closed, visitors need that information before they build their day around it. A zoo’s website, map, ticket confirmation and frontline team should use the same operational facts. That work sits close to membership communication: returning households notice whether the institution is helpful and candid.

A conversational visitor assistant belongs on the public side of this boundary. It can explain published programs and direct people to current access information. It should not improvise animal availability, health updates or veterinary explanations. The broader commercial uses are explored in our article on conversational AI for cultural attractions; the care systems discussed here have a different responsibility.

The next investment may connect existing work

My reading of these developments is that the most consequential change is likely to come from better connections among observations, individual histories and conservation resources. Each connection makes it possible to ask a more useful question. It also exposes gaps that a new interface alone cannot repair.

Before the next budget discussion, I would ask the animal care, records and education teams to bring one example of information they repeatedly have to reconstruct. Follow that example from its origin to the person who needs it. The resulting conversation may point to camera coverage, observer training, an accession workflow, a data agreement or a clearer public explanation. Those are very different investments, and the distinction matters.

For mixed institutions, the aquarium technology article follows a different path through water systems and conservation research. Our zoo, aquarium and wildlife attraction work connects these institutional strengths to a public proposition people can understand.

Make the work behind the visit easier to understand.

PaperBoat Media’s zoo consulting and advisory work helps leadership connect visitor communication, membership and growth to the institution’s actual capabilities. A useful starting point is the gap between what your team does and what your audience understands.

Research note: linked examples are published institutional or product documentation. The implementation judgments and hypothetical habitat example are editorial analysis.

Scroll to Top