Challenges

1. Aakash - What’s the content?

One of the motives to make Aakash available for students was to have the NPTEL courses accessible on them. Now, with the launch of the Sakshat Portal, can you think of ways to improve the standards of suitable educational content, and how they can be delivered directly to these tablets? Availability of MOOCs - similar to the availability of Coursera courses on the iPad, might be a first step. Is there another model? Feel free to create Native apps/web apps/mock apps that can run easily on an Android platform (Ice cream sandwich) that proves to be student friendly.

Debate Question: Is it time to rethink how we deliver content on these tablets?

Video

Pros:

  • Understandable
  • Colorful and engaging
  • Ability for the best teachers to really get across to students Issues and solutions

How to deliver video content to millions of tablets?

In house streaming is difficult since most places do not have high enough bandwidth to provide uninterrupted video to many students (3G coverage is very low, and GPRS is too slow).

So another option is to distribute videos via physical USB connection or wifi connection at schools. This might work better, since now the students will also have an incentive to go to school. A physical connection is also needed for another reason described later. We can compare these options now. We are assuming that there is a dire shortage of teachers and hence everybody has to get video for all subjects, and most of the learning happens through videos. Teachers are there just to resolve doubts as in the inverted classroom model.

USB

Four or five fair quality videos of 1 hours each can be ~ 2GB. Assuming these tablets are pretty low end, and can only get a 5 MB/s transfer speed, this transfer will take 400s = 6 minutes, which is not too bad. If we assume all the transfers will be done within a timeframe of 4 hour, and there are ~ 400 students, we will need ~ 10 usb outlets, which might not be too difficult to provide. We can share one usb port from a computer via usb hubs.

Wifi

Assume our low end tablet can again get wifi speeds up to 1 MB/s (802.11b is 11Mbps, which is similar). The videos now download in 40 minutes, which is reasonable enough. But, too many students downloading will definitely reduce the overall bandwidth to each student. Assume we have to provide the videos within a timeframe of 4 hours to all the students. Now, there needs to be at least 75 wifi outlets which will be much more difficult from a cost perspective.

Battery consumption

Videos consume lots of battery unless we have specialized video IP blocks within our processor. Even then, a low end tablet will not have more than 4 hours of battery life, which means our student will have to recharge in < 2 days if he is watching 2 hours of video everyday. Considering the poor penetration of electricity, probably providing usb outlets for charging at schools will be better. Assuming a tablet takes 1 hour to charge, this will require ~ 50 outlets per school (again assuming we have a 4 hour window in which anybody can charge their tablets).

Thus it seems video distribution needs a USB hub at every school containing ~ 20-50 outlets. One option to lower this is to provide one tablet every 4 or 5 student group, which will then lower the necessity for these outlets. Such groups will also spur group interaction, which can be a good side-effect. If USB is the way to go, it makes sense to build in more storage capacity so that students can refer to older material, which is a big advantage of online education. A 32GB model still can hold only last month's videos.

Audio + ebook

Pros:

  • Easier to distribute
  • Consumes less battery, thus overcoming the larger hurdles of video distribution
  • Can have fair amount of interactivity, if we include a pdf reader with each device, in audio questions can be done in the pdf.
  • Low storage requirements. Assuming audio is 1/10 of size of equivalent video, we are looking at a full year's audio + pdfs all in one 32GB device.

Issues and solutions:

  • Not interactive, difficult to show logical steps than video, but better than books, and can be nicely complemented by a book or pdf with pictures.
  • Fairly cheap devices can still have interactive text and audio which can again be downloaded via USB. Cheaper devices will also mean the government has to foot a lower bill.
  • It seems better for each student to have one such cheap device. This can be based on e-ink technology like kindles, which will then cost much less and last much longer.
  • Fairly sophisticated or specialized material can also be distributed via this method to interested students.
  • This model is fairly non-interactive within different schools. In order to remedy this, we can encourage students to participate in question answer communities through their common tablets. Maybe we can also provide an SMS-based interface to asking/answering questions through the kindle-like devices.

Challenge:

Build a cheap and efficient device for reading pdf’s and listening to audio lectures by using an e-ink display. One can use reused displays from kindles available from ebay for ~ 25-30$. A resistive touchscreen can cost ~ 10-20$. After having a proof of concept on a board such as raspberry pi, one can look into building a more self-contained device. Intuitive alternatives to a touchscreen (joystick/arrow buttons) can also be investigated, since alternative keyboards in various indian languages can be difficult to mass produce.

4. Networking of Campuses in India

Under the National Mission on Education through Info and Communication tech (NMEICT), each university and college is touted to get a broadband internet connection and standalone network-based solutions are being thought about for very low cost connectivity to student’s home. What are the ways to implement this (topology, cable type, server hosting capabilities, etc)? Can we extend this to schools, to build the primary internet access backbone for students?

Project Members

  • Aneesh Nainani
  • Pranav Pai - pranavpai (at] codeforindia [dot) org
  • Ninad Vartak - ninad.vartak [at] gmail [dot] com
  • Tushar Dadlani - tush726 [at] gmail [dot] com
  • Sagar Bhosale - sagarbhosale019 [at] gmail [dot] com