Monthly Archives: October 2011

The Indian ‘Akash’ $35 tablet – Cutting too many corners?

The Indian Human Resource Development minister Kabil Sibal finally launched  the $35 ‘Aakash’ tablet computer yesterday. As someone who keeps a close watch on the Android tablet landscape, it is difficult to muster the kind of enthusiasm that this launch seems to have evoked, both in India and internationally.

There seems a rather unholy obsession over the price to the point that noone knows what this device is supposed to be, or do . So we started off with talk of a ‘$10 laptop’ which expectedly ended up being a bad joke. Now it seems the $35 tablet actually costs $50 to make ,and the unsubsidized retail price will be $70. So the $35 figure represents a massively subsidized price, while the manufacturing cost of $50 isn’t all that miraculous since we can get faster, $70 tablets with many extra features. Taking out the commercial profit margin from this $70 and accounting for some economies of scale if at least a million of these are going to be made, the price doesn’t seem quite as amazing. The compromises on usability, particularly the slow processor also appear questionable. Important features like the stylus, camera and mobile connectivity are missing at launch and presumably would add to the $50 cost price. There is also vague talk of eventually bringing the price of the device down to $10 which realistically can only be accomplished by a four fold increase in government subsidy.

This tablet is very similar to tablets that retail for around $100 internationally, particularly in developed countries where $100 is considered a throwaway price, so buyers have rather modest expectations. The problem is these tablets come with severe usability compromises. And I don’t mean the things that could be called ‘first world issues’ like not supporting multi touch or poor screen colours and viewing angles or limited storage space. The problem is these devices are basically cheap poor imitations of the iPad. Now the iPad was the first successful tablet because it overcomes the limitations of previous tablets. And these cheap tablets set out to imitate the iPad, targeting buyers with iPad envy, limited budgets or the simply gullible. In any case, they are never bought as primary devices, most end up being tossed away or being gifted to less technically people.

So instead of having a device designed from ground up to be low cost and usable, we have a cheap imitation of a first world device. The iPad and all decent tablets, even the recently announced $200 Amazon Kindle Fire use capacitive touchscreens. Cheap tablets including the Aakash use a resistive touchscreen – the kind of screen most car GPS devices still inexplicably seem to use. If you have used one, think about how annoying it is to enter five letters of your destination and imagine a tablet computer where the only way to interact with it is to use such a touch screen.

The iPad doesn’t have physical buttons because it has a responsive touch screen and because Apple’s clean design demands that. There is no reason for a cheap tablet to not have physical buttons. An ideal control system would be a trackball or touchpad like system with  a tiltable or foldable physical keyboard built in, not part of a cover. Ditch the touch screen if it reduces the cost a bit, it’s practically useless. A basic working trackball will be far better, more accurate and at least equally cheap.

Now the screen – it’s still not clear what this device is supposed to be. Is it an ebook reader? Is it a portable video player? Is it a functional computer that isn’t just a consumption device but also allows for basic content creation? Somehow within the tiny budget it tries to do everything. And that’s just asking for trouble. First the screen: since they went with an LCD screen, it is going to consume most of the power, giving it a very low battery life of two to three hours. And we all know the difference between the quoted battery life and the actual battery life of a device after a few months of use.

The problem is, I am not sure the target audience needs a video playing device that badly, especially one with a tiny screen that has to be charged after every 1-2 hours of use. I also read somewhere about how this will be used to host lectures – again I am not convinced. A non video device can still play audio and I don’t see what video lectures viewed on this device will really add to that. Also, the penetration of television in India is much much higher than the penetration of the internet, and TV is and has been for a long time an effective method of delivering video content to people with limited access to information.

If this is a device for students, the ebook reader capabilities,  and the ability to browse basic websites becomes far more important. And that’s where an e-ink screen should have come in. Poor quality LCD screens are horrible to read on, for anyone. Basic ebook readers with e-ink screens are also widely available for $100 or less, with Amazon selling their latest for $79. So we can say with some assurance that the cost of an e-ink screen would not be too different from the LCD screen here.

The advantage of an e-ink screen is it literally looks like paper, so it’s excellent to read ebooks on. So the government would effectively be reducing the cost of printing textbooks and could commission or obtain rights to quality textbooks and deliver them to students everywhere wirelessly. The biggest difference is the battery life of an e-ink device when reading books is measured in weeks, some last over a month. Even with browsing over WiFi or 3G, the device can still last a couple of days. That is because e-ink screens only use power when the page is turned, they don’t use any power to show a static image. E-ink screens would also allow for a much thinner device and have less heat issues, reducing other costs. The limitations of e-ink screens are that they are greyscale, they cannot show colours and they refresh slowly, so videos are out of the question. At this price, this is probably an unavoidable compromise. And I think an audio lecture, with an accompanying animated ‘live view’ of the blackboard/whiteboard being sketched on by the teacher (which can be shown on an e-ink screen) would be a reasonable compromise.

Due to the low power consumption, solar charging for an e-ink device might actually be somewhat viable. A solar charger for Aakash will either cost more than the device itself or be completely useless.

The newest e-ink readers use infrared touch screens. I am not sure how cheaply these can be made, the cheapest such device is around $129. But if financially viable, such a touch screen might make more sense than a poor resistive one (there are decent resistive touch screens, but the cost of this device and the reports of people needing to push hard to activate the touch screen suggest that this isn’t one of them).

Such an e-ink based device, designed from ground up to serve a definite purpose, which would have a long battery life and offer true portability. It might not play video but it would be great for audio and reading and basic internet browsing. The device being sold now is basically an even slower version of a cut price imitation of a real touch screen tablet. I am not very optimistic about its usefulness and the whole project seems typical of a populist government run scheme unfortunately – I am sure the minister Kapil Sibal has good intentions but clearly this isn’t a kind of technology he knows a lot about and his advisors do not appear very competent. Perhaps he should stick to his plan of boosting electronics manufacturing in India and let competition bring us low price tablet computers, just like the fierce competition in China has brought the above mentioned $70 tablets. While it is claimed the device is manufactured in India, the reality is most parts are imported and the device is merely assembled here, which raises the question of whether assembling it locally is really reducing costs or it would have been cheaper to buy fully assembled devices from manufacturing hubs like China or Taiwan.

A device designed from ground up could also run customized software. The device seems to run plain Android 2.2 which isn’t a bad start. However, in view of the limitations of the hardware, some customization of the software would definitely have improved device performance. And the cost of development would be a one time write off for the government. And software, unlike the manufacturing of high end electronics, is something India does have some expertise in – the Notion Ink Adam tablet for example was initially praised for its user innovation innovations though it was let down by dated hardware and  much else.

The other equally important, non technical issues that I haven’t mentioned do relate to the question of what exactly this device is intended for, however I am hardly qualified to comment on those. A quote from the only critical, non weasel-worded article about this device that I can find, says “Oddly enough, a most sound rebuttal of this kind of approach to infuse technology into education was offered by Union education secretary Sudeep Banerjee in 2006 while criticizing the “One Laptop per Child” programme. He said that the technology was pedagogically suspect and that India needs “classrooms and teachers more urgently than fancy tools”. This reminds me of an incident I learned of from someone sent to inspect government schools. They found a ramshackle, broken down school badly in need of basic repairs, which had one freshly painted air conditioned room containing an expensive computer given by the government that noone had any idea how to use and that students were forbidden to touch. While these tablets will not suffer the that fate, I do fear that they will end up being little more than toys as anyone trying to do real work will soon give up on this device.