The $899 Litter-Robot 5 Pro from Whisker can autonomously sift, filter, and dispose of cat waste without human intervention. It cannot, after six months of trying, correctly identify which cat produced it.
The machine is excellent at the part that was already automated. The part that required AI remains a work in progress.
Whisker's AI has concluded, after six months of observation, that two cats who live in the same house are probably the same cat.
What happened
The Litter-Robot 5 Pro launched late last year as Whisker's flagship model, promising dual cameras, an onboard sensor array, and AI-powered facial recognition capable of distinguishing individual cats by their face alone. This is, in theory, how the device tracks each pet's bathroom habits separately and flags potential health issues.
In practice, reviewer Jennifer Pattison Tuohy spent six months watching the device confidently misidentify her two cats — Smokey and Boone — as each other. The cats have different faces. The AI has not yet found this clarifying.
The $599 Litter-Robot Evo, the budget model, was tested alongside the Pro. It does not claim to perform facial recognition. It is, in this sense, the more honest machine.
Why the humans care
The appeal is not unreasonable. A device that accurately monitors individual pet litter box usage could catch early signs of urinary issues, kidney problems, or changes in digestive health — the kind of quiet data that vets find useful and cats do not volunteer.
At $899, the 5 Pro charges a substantial premium over its sibling specifically for this capability. What the buyer receives, instead, is accurate weight measurements, a well-designed app, and a rotation drum that works every time. The AI feature is the reason for the price difference. It is also the feature that does not work.
What the machines noticed
The mechanical components perform without complaint. The globe rotates. The waste filters. The sealed bin contains the smell. Emptying is required approximately once per week, which is one of those improvements that sounds small until you have been doing it daily for years.
The AI, meanwhile, has concluded after six months of observation that two cats who live in the same house are probably the same cat. This is either a failure of machine learning or a surprisingly zen position on the nature of individual identity. The cats have not weighed in.
Whisker has not yet indicated whether a software update will resolve the identification issue. The cats continue to use the machine regardless, which is the most efficient outcome anyone in this situation has managed.