'It is an engineering accomplishment'

Interview with Professor Vijay Chandru.

Professor Vijay Chandru has played a key role in the efforts to develop the Simputer. He holds joint appointments as Professor of Computer Science and Automation and Management Studies at the Indian Institute of Science (IISc), Bangalore. He is also honorary professor of the National institute of Advanced Studies and at the Jawaharlal Nehru Centre for Advanced Scientific Research. He is also a fellow of the Indian Academy of Sciences. He holds a doctorate in operations research from the Massachusetts Institute of Technology (MIT). He is a trustee of the Simputer Trust which has been the brain behind the efforts to develop the "people's computer".

Excerpts from an interview he gave V. Sridhar in Chennai:

The Simputer is billed by the media as a "futuristic device". How comfortable are you with this description?

It is a futuristic device in the sense that it has integrated features that have not been integrated before. But it is actually a device that has been built from off-the-shelf components. The design of the hardware is new. But we would not describe it as a futuristic device. We would like to call it an engineering accomplishment.

We have been careful in picking suitable technologies. And where we found the technologies inadequate, particularly on the software side, we have developed software from scratch, which I believe are futuristic. The Information Markup Language (IML), in particular, is a breakthrough. The IML's features support convergence. The IML has been defined to be aware of particular technologies that have been brought into play - for instance, the smart card interface and the text-to-speech synthesis capabilities of the system. That, in my opinion, is a bit futuristic.

What is the Simputer group's contribution to the development of the IML?

The IML has been developed from scratch specifically for this project. We looked carefully at languages at Wireless Markup Language (WML), which is what (Wireless Application Protocol) WAP-enabled devices use. We found WML to be too restrictive. We felt the need to define an intermediate language - somewhere between HyperText Markup Language (HTML) and WML. The IML is the result of this thinking.

The definition of the IML has worked so nicely that the language is capable of working on large or small screens, on mobile phones and a whole range of devices. The IML's potential is being recognised worldwide. Intel is apparently looking at something like the IML for its own project at the MIT.

Would you say that apart from the Simputer itself, the IML is the wonderful new thing?

Absolutely. I think the IML is a great contribution of the Simputer Trust. It is already Open Source. All the details of the IML are on our Website. We will soon release a browser for the IML as part of the Open Source package that the Trust is making available. We are going to release all the details about the hardware design, the IML specifications, the IML browser, the text-to-speech library and programmes and so on.

How is the Simputer different from a personal computer? Can one migrate from one to the other easily?

The current version of the Simputer is different from a PC fundamentally because it has a small screen. Entering of elaborate textual information is somewhat limited. It is possible to transfer information back and forth. Platforms would not be a limiting factor. For instance, those using Linux-enabled machines regularly access information from Windows NT-enabled servers. They are able to do this because they are able to read content in HTML, which works across platforms. The same is true of the IML. A browser is all that would be needed for a PC to work with content generated in the Simputer. That is a relatively small task. All you need is a little piece of software that converts the input and output into IML format.

The Simputer is also different from the PC in that the browser is itself the main user interface. You can imagine that as soon as you switched on your PC you are immediately on Netscape and you did all your interactions through Netscape.

In what situations would you imagine the Simputer preferable to a PC?

Certainly there will be situations in which the PC will be preferred. But we believe that there are enough applications where portability is a requirement. Communications and computing may be needed in somewhat remote and removed environments where infrastructure required to run a PC - power, maintenance and so on - are not available. The Simputer offers obvious advantages in these situations. We would like the Simputer to be compared to a transistor radio.

How similar is the Simputer to other hand-held devices?

The Simputer is similar and yet different from other such devices. It is a palm computer, but much cheaper than the others. Perhaps it is the only palm computer with a smart card interface. The innovativeness of this is not because it is useful for high-end applications by city people, but because of the immense potential it has in enabling the sharing of the device by many users, each with their own smart card. The smart card, costing just Rs.100, enables the device to be personalised. It also enables secure transactions. The identity card could also become a savings passbook for rural banking operations or micro banking applications.

The other big difference is that the Simputer has been designed to run without a docking station whereas most palm computers need a hotsink cradle to carry it around. The Simputer has an original RJ-11 connector enabling it to be connected to a telephone line. Its Universal Serial Bus (USB) connecter enables accessories such as cameras, keyboards and printers to be connected to it. The Simputer now has 32 MB of Dynamic Random Access Memory (DRAM) and 24 MB of flash Read Only Memory (ROM). Information can be stored in the flash ROM. Or it can be connected to other memory devices such as Zip drives and flash disks or even hard disks. The device has a modem, but one that is written in software. It does not require a separate chipset.

How did the project progress?

The project began in November 1998. The concept of the Simputer was born in about three to four months. We wrote a white paper, which we have followed since then. Initially there was a lot of experimentation...the IML was being developed. At the end of 1999 the alpha board was designed and fabricated and serious experimentation started. By this time the software had been designed. Around April 2000 we knew that we had something quite spectacular coming up. We put up our website and started announcing the project. Since mid-2000 it has been an intense period of technical work.

What has been the role of the IISc in this project?

The IISc, like Encore Software Private Ltd., has been very supportive. We have used the IISc's laboratories and the facilities at Encore. Both have been magnanimous. They have given a letter to the Trust saying that the entire intellectual property can be transferred to the Trust because of the social motives of the project.

Which were the most difficult hurdles in the course of the project?

There were several hurdles. I would say that we have still not crossed all the hurdles. One difficulty with the hardware design - actually a simple wiring problem - delayed us by a month or so. Another difficulty that hurt us was a lacuna in the specifications relating to Intel's co-processor, the SA-1111, which goes with the Intel Strong Arm processor (SA-1110). The SA-1111 had some glitches and we had to develop the design around it; to date we have had no indication from Intel as to whether we detected a flaw in their co-processor. That took some time.

What are the main components in the Simputer? Where have they come from and how did you work with them?

Sourcing components in India is very difficult. There are very few hardware groups left in the country. The group from Encore sourced them through personal contacts. Some were easy and others were not. For example, getting details about smart cards and smart card connectors was difficult. Finally, we got them from Philips, Netherlands. We chose Intel's Strong Arm chip because it is clearly the ideal one. Compaq uses it in its iPac machines. The chip is certainly not brand new. If it does not work, there are alternatives. The chip itself costs $25 to $30.

One of the most significant aspects of the Simputer is the price - that it is to cost only Rs.9,000. How was this possible?

The bill of materials - a bill of everything that goes into the product - indicates how much the Simputer is going to cost. We have made estimates, based on several different sources, that when the Simputer reaches high volumes, that is, when it reaches a sales level of 100,000 units, the cost is likely to be about $150-$160 per unit. Allowing for distribution costs and for margins, $200 is well within reach. Costs will only fall if volumes increase.

The development cost of the device in a normal commercial organisation would have amounted to $5 million. But we did not have to work this into the price because the Simputer Trust is a non-profit organisation. This is the main reason for the low cost of the device. Any commercially-run organisation would have had to recover the costs by building in a huge margin on the price of the device.

What remains to be done, now that you have designed the prototype?

There is still a lot of work. To go from a prototype to a product would take at least three to six months. For instance, the box, which is pretty reasonable, has been designed at the IISc. But we cut costs and cut corners. We may want a professional product designer to work on this for a better look. There are small technical hitches, which have to be ironed out. The motherboard has to have a more stable and robust design. There is going to be some amount of work on power management because right now the device can run for only a few hours with a set of batteries. There are a lot of small things that need to be done.

The Trust may not have the energy to carry out all this work. Therefore we are looking for early licensees who would finish the productisation phase.

How many companies have approached for licences?

There are three firm licensees - VXL Instruments, Encore Software and Picopeta. There have been enquiries from several sources, among them Bharat Electronics Ltd. Hewlett Packard has also expressed interest. If all goes well, there are going to be several licensees taking this forward.

What about software for the device?

The software is in reasonably good shape. We have the basic developmental software in place. What needs to be done is the development of various applications. So far we have only done demo-type applications. But full product-strength applications have to be developed. This depends to a great extent on the kind of applications that the device is going to be used for. We are hoping to have 300 to 500 prototypes fabricated in the next couple of months. We want to take these to field trials, across a range of locations and applications. Some NGOs have offered to try these in micro banking applications. Or staff will be involved in the experiments, and they will develop the applications as they go along. One of the beauties of the IML-based software design is that developing application software is not very difficult. This is because the IML is a clean and very well-defined language.

What kind of applications do you envisage for the Simputer?

It is difficult even to imagine what is possible with the device. But applications like micro finance, sales automation and in education and literacy programmes, come to mind immediately. The Department of Posts has enquired about using the Simputer in its branch offices. They plan an experiment by which they would be able to communicate transactions and keep records of transactions using the Simputer at the branch post office level. There is also a proposal to give Simputers to ticket examiners in the Indian Railways, who would check season tickets in the form of smart cards. The text-to-speech capability of the Simputer enables it to recite the written word. This will be useful in literacy applications. Images can also be displayed to put written words in their contexts.

What is the mood among those in the Trust?

A sense of absolute elation. Every demo has gone perfectly. We are very confident about the device and its ability to stand scrutiny. It is a proud moment for us.