India Officially Launches Simputer

aravind writes "The Communications and IT Minister, Pramod Mahajan, has launched India's indigenously developed low-cost handheld Personal Computer -- Simputer -- at an IT and Communication expo, SMAU 2002, in Milan. A low-cost handheld PC on GNU/Linux working through a browser for international markup language IML, priced at Rs9000 (less than $200). 200Mhz StrongArm processor, 32MB DRAM, 24 MB flash, touchscreen, speakers, USB, text-to-speech, MP3 capability ... " Look here for some of the previous stories we've run on the Simputer.

Comments

... and you poke at it...

[Anonymous Coward]

From the FAQ:

Q: How do I enter text? Can I attach a keyboard?

A: There are two options on the base simputer for entering text: one is a soft keyboard, that can be brought up on the touch screen and you poke at it to enter one character at a time...

I guess they took it to the point that that even the description is simplified. :)

Re:... and you poke at it...

[LittleBigScript]

Well, it could be simpler.

Do they mean to poke at the Simputer? or something else...perhaps the person sitting next to you.

Stranger: Hey, stop that.
User: I was only trying to use my computer. It says to poke.
Stranger: You must use AOL, too, huh?
User: ...

Ahh, humor, it cures all forms of frustration.
Even the frustration of being unable to poke people.

Cheap, but is it enough?

[syphoon]

Its good to see low cost applications such as the simputer being made to help those who otherwise wouldn't have access to any such device. But I wonder, in a country as vast as India with 58 taught languages and 87 different language newspapers (http://www.abhishek.mybravenet.com/languages%20of %20india.htm), how effective would the speech recognition really be, especially when you take into account the lower literacy levels of the demograph its aimed at.

Re:Cheap, but is it enough?

[metlin]

IAAI - I Am An Indian
IAANLPR - I Am An Natural Language Processing Researcher

Ok, although India does have so many different languages, the majority of the people speak a countable few, maybe with subtle differences in dialects. In fact, only about 14 languages are recognized as official languages of India, and almost everybody can speak two or more Indian languages.

So, although the total figure may seem big, using just one language like Hindi would cover significant percent of the populace.

Also, there _is_ a lot of similarity between a lot of the languages, both in the written and the spoken forms. So developing a general prototype system and then expanding on it regionwise would not be as mammoth a task as it may seem.

For example, a lot of the South Indian languages sound similar, have similar sounding alphabets, with a few differences in grammar. The basic difference would come in smaller parts of the language set and may need certain prefixed lexicon modifications.

If these things are going to be custom built for each of these states, then I'm guessing that you'd have a system that is custom-built to the languages of that region.

It may take a while longer and maybe a little tedious, but I suppose that would be just worth the trouble, especially after having come this far.

Re:Cheap, but is it enough?

[The Cydonian]

Not to karma-whore, but just a few additions/corrections (I had posted earlier on the scripts; now including stuff on grammar as well here) :-

• The official 1961 Census listed 1652 "mother tongues".
• Another report suggests that there are 418 "listed languages"
• However, there are 18 (not 14) constitutionally recognised languages.
• Anthropologically speaking, there's little similarity between *most* languages except for geographical proximity; for instance, folks speak Gujarati, Siddhi and Indo-Portuguese in and around the islands of Daman and Diu. It's interesting to note that these languages have derived themselves from Aryan, African and European roots respectively. Very little overlap, historically speaking, but these languages are spoken nowhere else, hence are uniquely Indian.
• But despite all this, there *is* a certain amount of overlap. There are 10 uniquely Indian (as opposed to Indic; Indic would include scripts such as Burmese, Thai, Tagalog and Sinhala as well) scripts which are all derived from the 5th century Brahmi script. A fact that was already researched and recorded in C-DAC's masterful ISCII (Indian Standard Code for Information Interchange) format, which, as I understand, treats each individual language as a specialisation of an existing meta layout.
• It is my contention that the basic grammar for *most* prominent constitutionally recognised languages is inherently same; the grammatical difference between, say, Hindi, Telugu, Malayalam, Marathi, Urdu and Kannada is noticeable, but not significant.

Bottomline: I'm an NLP researcher myself fascinated by languages (see my sig). As much as I'm excited by this project, I really think we shouldn't kid ourselves, coz:-

• Let's admit it; computerisation of *all* Indian languages won't happen in our lifetimes. Denying that would be to deny India's mind-boggling linguistic diversity.
• If you want to increase literacy in India, get your basics right:- increase the number of schools and increase their quality. Don't search for magic bullets. They won't deliver, even if they're tech-y stuff.

Re:Cheap, but is it enough?

[metlin]

Interesting post! :-)

I'd however, like to add a few points which I consider relevant :

I absolutely agree that a lot of Indian languages have little or no relation to each other, but the truth is that a significant percentage of these languages can be traced back to a common root. For example, certain Brahmi-derived languages and Prakrit have very common grammatical rules, but are also very different in other ways.

As long as we can find a set of common languages which would serve the majority of the populace, it'd be great. Reaching that in itself would be significant achievement.

Regarding the differences in grammar, yes, you're correct. But look at it from the point of voice recognition. Malayalam and Tamil sound very much alike, and a person with the knowledge of one with absolutely no knowledge of the other can actually understand the other one. But the alphabets of Tamil and Malayalam are phonetically miles away, Malayalam has a alphabets that spell like Brahmi, look like Tamil ones and the lexicon has a lot of Sanskrit. Likewise for Telugu and Kannada. In fact, Kannada has grammatical rules that are pretty much like that of Sanskrit (there is a book called NLP - A Panian Perspective that discusses exactly these issues).

Well, as for what you said,

Let's admit it; computerisation of *all* Indian languages won't happen in our lifetimes. Denying that would be to deny India's mind-boggling linguistic diversity.

Perhaps not all of it, but once you have a large chunk of it, you'll realise that a lot of them are evolved dialects and can be traced to a series of common roots. Look at Urdu, Arabic script with Hindi, Persian and Arabic words in the North, while down south you have a mix of Marathi and Telugu words. So it may not be all that impossible.

Let me rephrase your statement - *Complete* computerization of all the languages will not happen, but basic computerization might just happen, however mind boggling that task may seem.

If you want to increase literacy in India, get your basics right:- increase the number of schools and increase their quality. Don't search for magic bullets. They won't deliver, even if they're tech-y stuff.

Hmmm.. I think the Simputer was originally intended more as a tool to help the farmers and the rural people, not to educate people. In fact, I fully agree with you that technology will not be the only saviour. But then again, tools equip people better. Don't look at the Simputer as the end result, look at it as a tool that'll ease your way into achieving it :-)

Average montly salaries

[Drunken Coward]

The average monthly salary in India is somewhere along the lines of $37. A person earning that much could hardly afford the luxuries of such a handheld, even if it could be attained for the paltry sum of $200. Pennies to us, but to them it could take a lifetime to acquire that amount of savings. Until we work to attain far cheaper methods of building computers, these people will be unable to experience the very joy we take for granted. Of course, this is a noble effort and no doubt will further number of Indians able to participate in the IT field, but further effort is needed.

Re:Average montly salaries

[rovingeyes]

Well, even though the average monthly salary might be low, you are forgetting the point that even if 2% of more than a billion population plan to buy (believe me there are people india who are filthy rich) it'd be lot more than the simputer guys could hope for. Certainly you can look at it this way - IT IS A START!

Re:Average montly salaries

[Albanach]

By reading the FAQ you would learn that the simputer has been designed for sharing by communities. THe examples they give are schools and community centres which are already established as places where shared resources are available to comunities.

The simputer is equipped with a smart card reader which is intended to provide personalisation to the device. The aim is to reduce the cost of _access_ (that's the important bit, not ownership) to the device to that of owning a smart card, not of owning the device itself.

Think of being able to walk into a local library and borrow a computer for a day instead of a book.

Re:Average montly salaries

[arvindn]

Pennies to us, but to them it could take a lifetime to acquire that amount of savings. I live in India, and I what you say is simply not true. You're missing the point that not everyone earns close to the average salary. An Indian "middle class" family (such as mine) could easily have an income of Rs. 50000 ($1000) per month. (That's > 25 times your average!) True, the percent of affluent people is small, but in a country of a billion people its still a large number. I don't mean to troll, but slashdot seems to have the general attitude that India is still a country of snake charmers and tightrope walkers. Get over it. While a majority of the population lives in poverty (and consistently gets screwed by self-serving governments), there is also a large, educated, wealthy technically minded workforce.

Actually, the reason the simputer took so long to take off was that its creators initially focussed on the wrong market - the illiterate masses. No company came forward to mass produce it and only the intervention of the government saved it from dying out altogether. But now that it has gotten off the mark, I think there is a very good market for it out here.

Re:Average montly salaries

[fault0]

Yes, but a larger group of people might pool their resources and get one. I've heard of urban poor (note: not villagers) buying TV's/with Cable in India this way.

In India, less than USD 200 IS A LOT !

[Taco Cowboy]

USD 200 might sounds peanuts to many. But in India, it's a hell lot for the majority of Indians.

Simputer is a good concept, but with RP 9000 price tag, I think it'll only benefit (again!) the Rich, the Powerful and the Corrupted, in India, of course.

Can anyone here offer any suggestion as to how to lower the cost ?

Thanks in advance !

Re:In India, less than USD 200 IS A LOT !

[nomadic]

But it's not like India doesn't have a middle class. From the posts here you'd think everyone's either a pauper or a millionaire; there are millions of Indians for whom $200 USD would be an investment, but a doable one.

IT doesn't replace education.

[Krapangor]

This is a really stupid waste of money.
India's problem are not people who don't know how to use a computer, India's problem are people who can't read or write at all.
What use whould such people have for a computer ?
I doubt seriously that it had Hindi speech recognition (Hindi is much harder to do than French or English).
So these people would be able to buy for a 2 years wages a high-tech doorstopper.
That's classic wasted goverment effort. How about building schools instead ? Or creating decent taxes to distribute the enourmous wealth of the rich to the poor one so that they can efford education or even a real computer in some time ?
But as always technology without meaning.

Re:IT doesn't replace education.

[vsync64]

What use whould such people have for a computer ? [...] So these people would be able to buy for a 2 years wages a high-tech doorstopper.

Don't underestimate people, even poor brown people.

Re:Posted @05:14AM EDT?

[Twirlip of the Mists]

Let's see how many days it takes for CmdrTaco to notice the problem and fix his cock.

Funniest typo ever.

from the FAQ

[jas79]

Q: Can I create a Beowulf cluster using many Simputers?

A: You must be a /.er; in which case you know the answer!

Re:from the FAQ

[Large Green Mallard]

I thought you were just making this up.. but this is actually on the FAQ.. how cool :D

Re:from the FAQ

[silentbozo]

Given that there's a USB port and a phone port (powered by some sort of software modem), you have two choices to network your equipment. The real problem with clustering will be power -you'll have to hack each case with an adapter to supply 4.5vdc. A more elegant way of handling this would be if they allowed you to supply power to the simputer via the USB port (say, could you get enough current to charge your batteries?)

It's funny that they talk about client-server processes for the simputer, since it lacks wireless access (when I think of portable devices talking to other devices, I think wireless.) However, if you can implement a common interface for connecting to a network and charging via a common port (could USB work?), you could install ports all over the place.

With this kind of distributed computing in place, India could soon be home of some serious computational power...

A few things about India

[panurge]

First, India has a big educated middle class. Over 100 million people. They don't have the sort of incomes we do in the West, but a $200 handheld is within their possible budgets.

Second, India has huge potential in IT as their materials-poor economy has encouraged education in mathematics and other subjects which do not require expensive learning facilities - you do not actually need a computer to learn computer science, but it sure helps.

Third, India cannot afford lots of imports from the US, Korea or Japan. They need to be self-sufficient (even if it wasn't one of Gandhi's principles).

Fourth, the demand for such things is enormous. Believe me, I once thought I was going to find myself in prison in Mumbai because I had an HP calculator and a mini circuit tester in my luggage ("Admit, you have brought these to sell on black market")

Fifth, even poor Indian villages have the odd educated person who will provide services for the locals - and such people would benefit enormously from a handheld. The idea that every peasant should ultimately have a compactflash/smartmedia card with all their own information on it, is actually a hugely enabling one in a subliterate culture because it allows them access to a personal store of information. If it has to be retrieved by symbols on a soft keyboard and text-to-speech, does it matter?

Unfortunately, looking at some earlier posts, India and China are far from having a monopoly on illiterate peasants who don't know what goes on in the rest of the world (flamebait)

Available outside India?

[Ed Avis]

So are these things going to be available in other countries? Assuming the software is available (and it's Linux, so why not) then this thing could stomp on the Ipaq and other more expensive handhelds. At least for price-conscious buyers such as schools (the old Psion Series 3 and 3a had some success in British schools marketed as the Acorn Pocket Book - and it's a lot cheaper to buy ten of these handhelds for a classroom than a couple of PCs).

Simputers need not be for an individual

[Anonymous Coward]

I had attended a talk by Vijay Chandru (the no. 1 contributor to it) & they said their aim was not only to provide individuals with cheaper comps. but to provides groups of people as well.

In India, many village children (as well as grown-ups BTW) have never heard of comps. & even in cities. not many schools (incl. mine) have more than 10 comps. Those schools can instead invest on 10 simputers (for probably 1000 students!) which would be more cost-effective.

They also say they don't want hi-fi speech synthesiser/recogniser as to learn a language (which is what village students as well as other villagers are expected to do), that's not required. It's OK if there is no proper intonation. The villagers can probably learn intonation later on but learning to write/read something even in their native language is still a great breakthrough.

The major problem faced by them is discontinuation of StrongArm processors by Intel. It's obviously very expensive to design a processor for simputer in India today.

Re:Speaking of reliable power....

[silentbozo]

If you're a geek (probably, if you get one of these), you'll probably also find a way to get your hands on some rechargable nicad batteries, and a charger (solar or conventional) to charge em with.

For those who get one of these gizmos, who aren't geeks, they'll just find a local geek to hook them up with the requisite technology. I expect the local village tech to build a side business supply them with a set of rechargable batteries, which he'll recharge at his shop for a fee, if there doesn't already exist a service like this now...

Finally

[abhikhurana]

So finally the simputer has been launched. But that doesn't mean it works flawlessly. Until about 2 months ago, they had some problems with their text to speech software, dhvani, and they were planning to go for a closed source solution from another indian company.
Now for some of teh concerns raised in the article, as usual about average India salary and stuff like that. The aim of the project is not that every peasent should own a Simputer of his own. In India, in villages u have small committees which are elceted by the villagers, and these committes are allocated some budget by the government. So the idea is that every committee buys one Simputer and then the villagers can simply use thier own compact flash card if atall required. I mean its use was forseen in making weather prediction services and agricultural help availaible to the farmers. For that you dont need ur compact flash. If you want advanced services, then u can buy one, but then compact flash cards are not all that expensive.
As far as the Indian middle class is concerned, currently they account for around 40% of the population. Not all of them will feel the need of buying something like this, but they very well can. That is 400 million people.
I am just waiting for a review now. Maybe I will get one in December and write one myself. I enjoy working on stronARM and I think it will be interesting to hack this thing.

A couple things

[gralem]

First: IML stands for "Information" ML, not "International" ML.

Second: The product is not shipping, it was just presented by the IT Minister of India. No shipping date has been set by any company. Aparently the people at simputer.org do not build the product, they licence the hardware to be built. There are no listed manufacturers of the simputer.

It is not shipping, it is not available. (But according to the FAQ, it should be shipping by March 2002!) All said, the hinduonnet article is simple marketing fluff (ala M$, RH, etc).

---gralem

The features do not really matter...

[jukal]

Although the featureload of the Simputer is quite convinving, the main point is that this is now obviously the first truly open hardware project to have actually entered the martket. Let's see if it opens the floodgates....In the future, however, the featureload might be even more impressive as the open HW approach surely enables fast and cost-effective development.

Form factor

[e_n_d_o]

PCs can get into the $200 price range. Systems can be found for as little as $200 at Walmart (sans $99 15" monitor).

Based on what I've seen here, I imagine it would have been possible to develop a system in the $200 hardware price range with a 13" monitor. I'm curious to know why they didn't choose a larger form factor for the machine. The advantages of the PDA-style design are portability, power consumption, and a pen-based interface. The cost is a tremendous restriction in capability, and the requirement of developing properietary hardware. I imagine that portablility will also often be a negative, as the device is a handheld and its a fact of life that people drop things (of course, I'd be much less likely to drop my PDA if it cost me a year's salary).

These devices sound like a remarkable achievement, and I wish them nothing but success. But I am curious as to why they didn't go with a bit bigger of a box.

Re:Form factor

[Atrahasis]

The three advantages you list are exactly what they need. 1/ Low power consumption - uses batteries, so you don't need to plug it in (you'd be lucky to find somewhere to plug it in in the undeveloped areas, wouldn't you). 2/ Portability - so it can be shared by communities - is one of the main aims of the project. 3/ Pen based interface - so people can be taught to read and write using it. The two disadvantages : 1/ Limit in capability - less than 5 years ago my desktop was based on the StrongArm, and I can tell you it wasn't 200Mhz. It was still a lot faster than the ix86's around at the time. 2/ Requirement to build prop. hardware - the project doesn't manufacture - it licenses (or will) the manufacture out to tech companies. I'd say this was an advantage because they can keep the price down by using the lowest bidding contractor.

Please tell that to Malaysia

[Taco Cowboy]

Your message's title "IT doesn't replace education" is indeed very striking, considering that the government of Malaysia - a fifth-rate country, mind you - is on the path of REPLACING education with IT.

Specifically, the Prime Minister of Malaysia has commented on several accounts that,

"The teachers are mere facilitators. The main point
is the computers, where the students learn from. The
teachers just _help_ out if any problem arises."

Dunno what will happen to the children in fifth rate countries such as Malaysia.

Simputer/Literacy

[Omkar]

An issue that's been raised here is the uselessness of the Simputer to an illiterate population. Couldn't the Simputer use its text/speech capabilities to teach people to read? This would eliminate two huge problems with one stroke.

If you wanna buy one...

[notb4dinner]

... one of the licensees listed on simputer.org is supposededly making some evaluation versions avaliable soon. See here.

The real change...

[Cheese Cracker]

... will be when China has ironed out all the bugs with their CPUs and software. Then you can forget
Intel and Microsoft having much sales in Asia and the rest of the developing countries. A guess would
be that a Chinese handheld would go for $50. China has the ambition of taking the lead in the IT
market in Asia and the developing countries... and I bet they will. And then slowly they'll move over
to take market shares in the developed countries... maybe with 'inferior' products, but it will all go the
way the car industry went. Once they get a foothold, they'll make better and better products and finally
pass companies like Intel and Microsoft.

keep making fun of India REALITY CHECK

[Meorah]

After all, that's where all your jobs are going... might as well have a few laughs before you find yourself back in college either getting a masters in CS or another BS in Mgmt or Finance.

http://www.bigates.com/html/Pdf's/Benefits%20of%20 the%20company%20you%20keep.pdf

I can't believe how many of you goofs actually think India is some 3rd world country. Just because they have a very large proportion of farmers and field labor who live off their own sweat and blood doesn't mean there aren't a ton of wealthy people, especially in the cities.

In addition, did any of you einsteins think that perhaps they'll be selling this device in Europe (where it was unveiled), Asia, or god-forbid, N. America? Yeah, I'm sure all the poor people in the good ole U S of A will have to take out a 3rd mortgage to get their hands on one of these badboys... what with spending their life savings building beowulf clusters and all...

The lady on the simputer screen

[abhikhurana]

In case u guyz are wondering who is the female on the screen of the simputer
on simputer.org, well her name is Aishawarya
Rai. Beautiful lady indeed. The link also has her phone number but try at ur
own risk. Here are some
nice pictures of her.

Smartcard is $2, plus fees for access

[Bubblehead]

Software Development Magazine covers the Simputer in their "Deadline" section (unfortunately, the section is only in the print edition, not online). From the article: "For $2 and a nominal rental fee, each villager can buy a smart card that stores all his or her information, and allows Internet and e-mail access." Sounds quite feasible to me.

Re:So where can I buy one?

[Syre]

Look here.

Not for general sale yet, but you might convince them to let you have an Evaluation unit...