The before picture
The library’s circulation desk handled roughly four hundred patron interactions a day. Of those, an internal time-and-motion study found that two-thirds were routine loans, returns, reservations or reading-room access checks — transactions the patron could finish themselves if the interface let them.
The other third needed a librarian: reference questions, account holds, inter-library loan negotiations, payment disputes, accessibility help. Those interactions deserved a librarian’s full attention. They were not getting it, because the line for routine loans never cleared.
The kiosk did not replace the librarian. It cleared the queue between the patron with a real question and the librarian who could answer it.
From the post-deployment reviewWhat the kiosk actually does
Atena is a self-service touchscreen station, deployed near the entrance and inside reading rooms. A patron can:
- Borrow and return any item the library circulates, with the RFID stack the library already had.
- Reserve and pick up reservations from a hold shelf, with one tap to confirm.
- Pay overdue fines by card terminal embedded in the kiosk — no separate cashier interaction.
- Print, scan and copy from the same identity, with quota deducted from the same account.
- Reach a librarian through a clearly-labelled help button when the workflow does not fit what the patron needs.
What it doesn’t do
A short list of features we considered and refused:
- No “recommended for you”. The kiosk is a transaction surface, not a discovery one. Discovery happens in the catalogue.
- No upsell. No prompts to sign up for newsletters, donate, or rate the service. Each one costs trust.
- No identity sharing. The kiosk authenticates the patron, performs the transaction, logs out. The next person sees a clean welcome screen.
The reservation paradox
A finding the library did not expect: after the kiosk went live, the reservation no-show rate dropped from eleven percent to four. The team had assumed self-service would make reservations more casual — easier to make, easier to forget about. The reverse happened.
Our reading: the friction of asking a librarian to make a reservation was the same friction that made the patron treat it casually. The kiosk made reservations feel like a commitment the patron had made themselves, because they had.
Where the saved hours went
The library tracked what librarians did with the hours the kiosk freed. The largest single category was collection development — reviewing acquisition requests, weeding low-use materials, curating reading lists with the faculty. The second was reference depth: longer conversations with the patrons who actually needed them.
The third was visible. The Director of Libraries told us, somewhat dryly, that the team had used some of the saved time to argue more passionately about which serials to renew. Twelve months in, that argument was producing better decisions.
