Automatic Book Scanner To Bring Knowledge To Ethiopian Students

In 2013 the dean of an Ethiopian university addressed Maker Faire Hannover and outlined one of his concerns; that the high price of developed-world textbooks was holding back the cause of education for universities such as his own in developing countries. He was there to ask for help from the maker community to solve his problem, and a group of his audience took up the challenge to create an affordable and accessible automatic book scanner.

Their scanner builds on the work of Google engineer [Dany Qumsiyeh], whose open source linear book scanner turns pages by traversing the opened book over a triangular prismic former such that pages are turned by vacuum as they pass over carefully designed slots in its surface. Their modification replaces the vacuum with the Coandă effect, to more gently tease open each page and it is hoped reduce the chance of damaging the volumes being scanned.

The whole machine is controlled by a Raspberry Pi, and the scanning is performed by linear scanning optics, sensors, and electronics taken from flatbed scanners.

An important design goal of the project was to ensure that the scanner could be built without special tools or expertise that might be difficult to find in a developing country, as well as that it should be as inexpensive as possible. The frame of the machine is off-the-shelf extruded aluminium, and the body is acrylic sheet which can be cut to shape with a hand saw if necessary. It is estimated that the device will cost in the region of 500 Euros (about $568) to build.

More information can be found at the project’s web site (German language, Google translate link), including a selection of videos such as the one below the break showing the device in operation.

Writing from the perspective of having been peripherally involved in a professional book scanning operation at a large publisher the benefits of this machine are immediately apparent. Removing the binding and automatically scanning each page as an individual sheet produces a very fast and high quality result, but by its very nature damages the volume being scanned. This machine promises to deliver a solution to the problem of book scanning that is considerably less intrusive.

It is also worth noting that the project does not address any copyright issues that might arise from scanning commercially published textbooks, though this is more of a concern for the end user in terms of what they scan with it than it is for the maker.

Book scanners have featured many times here on Hackaday. We covered Google’s release of the linear page-turning design, and we’ve covered a number of page-turning designs as well. It is sometimes surprising how much of the world’s knowledge still resides only on the printed page – a lot of publishers don’t even have proper electronic texts of titles still in print, for example – so it is likely that this is a field with plenty of scope for further innovation.

The HackadayPrize2016 is Sponsored by:

35 thoughts on “Automatic Book Scanner To Bring Knowledge To Ethiopian Students

  1. textbooks aren’t expensive because of production costs.
    if publishers wanted 3rd world countries to read their textbooks for free they would just provide the electronic copy.

    1. I think that’s the problem, since they won’t give the electronic copy they acquire one hard copy and make it possible to electronically copy them. That would be illegal but only in countries where there is a copyright law agreement and a working law system.
      Another variation is to use very old textbooks from before the electronic age, ones where there is no copyright on anymore, maybe because the publisher long ago went out of business.
      And a third possibility is to scan textbooks from other than western/capitalist countries that perhaps did not publish electronically and once you have the electronic copy run it through an automated translator the distribute that translation.

  2. The Internet Age means information is FREE and now everywhere. I don’t understand why schools are still seen as normal, and a way to continue. Education software is getting to be one huge pork barrel! Testing sessions crash, students aren’t happy guess who else? Buggy makers still make a few in northern Indiana and Lancaster County Pennsylvania, but the rest drive petrol for now. I can’t see developing places not being able to have the human resources to hand scan books for common access. It’s something you will only do once.
    As anyone has felt turning a…just one…crump there I got it… page it’s not simple.

      1. He’s comparing not having free information this day and age and having the system of expensive textbooks to riding a horse and buggy instead of a car, like is done in Amish area (and why they still make buggies there). And saying it’s outdated to not have information available for free or very low cost.
        He’s arguing also that you can scan by hand because it only needs to be done once, and obviously in the very poor countries getting a person to do it should not be that much of an issue so I see his point there.

  3. Hmm. to decrease nois leven, I would suggest using a refrigerator pump and a small nozzle with a electric water valve from a washing machine. a small puff on the right moment should be enough.

    1. That seems in line with what you might get in the developing world from scrapheaps, so it’s not a bad idea.
      Although using old car parts might be even more available worldwide. But I’m not familiar enough with the details to guess what could be used from a car to get equivalent functionality.
      I don’t think those extruded aluminium profiles are that common in the developing world though, so I think there too it might be better to use used parts to reduce cost and increase availability. Because $500 is quite a lot in many places, so reducing that to 50% would make a big impact.

      1. Aluminium profiles are not produced in Ethiopia (paper isn’t also). I guess most aluminium profiles are produced in China but they are used in some other countries than China as well. In our latest design we are using 15 meters of this profiles. One meter cost about 3 Euro (already cut to the precise lenght you need)

      2. We also looked into the option to build the Scanner from other materials – wooden frame with HDF wood fiber plates – but we learned that it is much harter to get everything aligned. That way it would be a lottery if the scanner would work as desired. Our number one goal was that everybody can build this scanner w/o special tooling or knowledge and have a reliable machine in the end.

  4. So the way to get around the cost of high textbooks is to circumvente copyright protection and make a scanner? Huh? Alot of things cost alot. So the solution is to make an automated way to scan and copy such books. Their solution is to make a way to pirate said material easier?-

    1. That is what it is. A nicer ( at least, more honest ) way would be to that so-called dean to contact publishers and book dealers, explain his problem, and ask for donations or price reduction in used books, old editions, etc.
      I know, books cost a lot in my country also. But if we want to pirate one ( “xerox” it ) , at least we need to pay for the copies´ cost. If someplace has the means to read electronic books ( tablets, cell phones, internet , etc ) they can scrounge the funds for buying some.

    2. If they want english language textbooks then they can find a lot of them online already, no strong need to make another scanner. If they want texts in one of the languages of ethiopia then what is possible depends on ethiopian copyright law and how its fair use clauses relates to academic activities (teacher and/or student use). Which I’m sure you are an expert on and will now proceed to give more detail on, as an explanatory background to your earlier comment.

  5. It’s the publisher that is the crook. Book costs are high because the publisher wants 98% and the author gets 2%. When writers publish their own electronic books (Kindle) the costs are reasonable for everyone. The publisher cries because they have to spend so much on marketing (which they don’t really do), etc. Cry me a river.

  6. I dunno, textbooks are often quite thick with lots of very specific graphics, illustrations & images… I’m guessing they are printed in runs of thousands, not tens or hundreds of thousands. that gets expensive, particularly when they have to excise chapters on evolution & human sexuality for different states. Printing books is not the same as printing business cards. I suppose the most efficient hack would be to crack the publisher’s servers and grab the PDFs, but that would be no fun.

    What might the reaction be if this were a hack to convert a juke box into a machine that would take a stack of 45s and convert them to MP3s? … golly, everyone is off-topic

      1. We must keep in mind that in the developing world they might not have the industry and availability of print shops that we have, and furthermore you must also keep in mind that in many instances textbooks have to be in local languages to be understood.

    1. Our approach is not to do any illegal. For the purpose of teaching and to be used within an university context it is legal to copy or digitize books in Ethiopia (and Germany and I guess a lot more countries).

  7. That’s a much better design than the usual book scanners where they fold it open and flat and photograph the pages.
    Assuming they could produce these, this could be useful for places scanning in books (Google, Gutenburg, &c) and cause less stress to old books by not opening them as much.

  8. Just a hint for serious book scanning – back the pages to be scanned with a black backing. This eliminates the bleed-through reflection from print on the reverse side of the page. Most scanners use white backing because they expect the pages to go to a printer and white eliminates the chance for a black border that few printers do well, but it does so by increasing the reflection from light passing through the page.

    I found that scanning cleanup effort was nearly eliminated using black backing.

    The problem in two-face scanning, like this, is getting the black backing behind a page that hasn’t been turned yet, but if the book had only single face scanning that could be done; just run the book through twice – once for odd pages, the second for even pages. Or de-spine the book (aaaaahhhhhh!!!) and scan the pages in order with a page flipper to get the second side.

Leave a Reply to KarstenCancel reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.