The biggest issue with scanning books is not the scanning itself, but instead dealing with copyright law. As mentioned in the article “What about copyright law?” there are some hurtles to overcome in order to get complete access to all the worlds books, and even thought a library may own a physical copy they may not have the authority to scan that book.
There are 2 ways that libraries can get access to digital books.
First libraries can get digitally scanned books (that is if they are available) from distributors like Overdrive or 3m Cloud Library, and e-books are also available from other sources including Amazon or The Digital Public Library of America. Some books are free and others have usage fees the library has to pay.
Second, libraries can digitize their own books; as books are requested by library users they can be scanned and made available in digital form.
In recent years the technology to perform the actually scanning of books has improved dramatically. There are some book scanners that can now scan over 250 pages a minute, like this one made by BSF-Auto; Video BSF book scanner in action.
The key would be a completely automated, mobile robot, that could receive a request from a library user, determine if the book is available in digital form, and if it is not, check the libraries’ physical inventory, then locate the book, take it off the shelf, scan it, then put it back delivering a digital copy of the book to the user that requested it. This approach would mean the library would not need additional staff to scan books; also requests could be made and fulfilled even when the library is closed. The technology for this type of robot already exists; it just needs to be put together and realized.