Computer systems have helped catalogue libraries for decades, but if some reckless reader has put a book back in the wrong spot, it's a daunting task for librarians to search the entire building for it – but not for robotic librarians. Researchers at A*STAR's Institute for Infocomm Research are designing robots that can self-navigate through libraries at night, scanning spines and shelves to report back on missing or out-of-place books.

This autonomous robotic shelf-scanning (AuRoSS) platform scans RFID tags on the books and produces a report. In the morning, the human librarians can check the results and can easily see which books are in the wrong spot and where they belong. There's still a need for human labor, but it's far less time-consuming than manually searching every shelf for misplaced titles.

The wheeled robot uses lasers and ultrasonic sensors to guide it through the stacks with precision down to the centimeter. "We decided to detect the shelf surface itself, and use that as a reference to plan the paths," says Renjun Li, one of the researchers on the project.

It does this with the help of a robot arm that adjusts the RFID antenna to keep within the optimal tag-scanning distance: too far back and the signal is lost, but too close and it will hit the shelf.

AuRoSS has been trialled in libraries in Singapore, where it achieved up to 99 percent accuracy, even with curved shelves. The researchers say that the system can easily incorporate different sensors beyond RFID, including cameras, Bluetooth and Wi-Fi, and could be adapted for use in warehouses, retail stores, or in the MedTech, Aerospace and Automotive industries.

Source: A*STAR Research [1], [2]