Comments for Professor Sadagopan's Weblog http://ss.emergic.org Views on various aspects of IT and Education from an Indian Perspective Sat, 11 Jul 2009 10:44:09 +0000 http://wordpress.com/ hourly 1 Comment on Farewell to Nandan Nilekani on July 9, 2009 by Dr. Vijay Jha http://ss.emergic.org/2009/07/09/farewell-to-nandan-nilekani-on-july-9-2009/#comment-19051 Dr. Vijay Jha Sat, 11 Jul 2009 10:44:09 +0000 http://ss.emergic.org/?p=1146#comment-19051 I wish him all success. However, government's intentions are doubtful, until it puts the framework of this project with its feasibility report on its website, so that the public can find out where and how the money is spent, and whether there are major loopholes in its feasibility. Reasons follows: Unique ID Card: A 150,000 cr or Rs. 10,000 cr Project On 29 June 2009 the PM announced the Unique ID Card project, with an estimate of Rs. 150,000 cr. In Sep-Oct 2000, the then Prime Minister had announced the same project under the banner of “National ID Card”, with an estimate of Rs. 6000 cr. On 26 June 2009, certain sections of the media (Like Siliconindia.com) had reported that this project had the estimate of Rs. 10,000 cr. In April 2009, the media had reported on the same project after talking to bidders of the ongoing PAN smartcard project, in which each card was supposed to cost Rs. 67. The latter card is similar to the UID Card except little inexpensive memory overheads. The centralized citizen’s date bank and related software for type-II query was also contained in the year 2000 project. So, the potential bidders, who have been in touch with the nodal agency – the NIC, had correctly reported the project estimate Rs. 10,000 cr. The framework of year 2000 project had suggested a pilot in at least a city of size Delhi. The then PM Vajpayee had announced for pilot tests in 8 cities. Till date the government has not conducted any pilot in any sizable city or state. So, how could the government be confident of the estimate or the feasibility of the project? Small pilots aren’t enough; success of small-scale typing jobs was no proof of large-scale failure of 1994-95 voter-card project with more than 20% errors, as reported by the then acting CEC Krishnamurthy to the TOI in the autumn of 1996. The year 2000 project was based upon a framework developed by the NIC, and the MHA (Home ministry), and its feasibility report was obtained from an Indian IT giant. Then the government had introduced the bill to the parliament. However, in Oct 2000 the Home ministry got convincing feedback that it was neither financially viable, nor legally nor technologically feasible, and so the bill, and the project were put into the cold storage. The NIC has been providing wisdom to the government on similar issues since the disastrous 1994-95 voter card project, in which Rs. 1000 cr was wasted. Several PSUs were contractors in many states then, and they also had more than 20% errors in 1994-95 voter card project. So, the government should put the framework with its feasibility report on its website, so that the public can find out where and how the money is spent, and whether there are major loopholes in its feasibility. I wish him all success. However, government’s intentions are doubtful, until it puts the framework of this project with its feasibility report on its website, so that the public can find out where and how the money is spent, and whether there are major loopholes in its feasibility. Reasons follows:

Unique ID Card: A 150,000 cr or Rs. 10,000 cr Project

On 29 June 2009 the PM announced the Unique ID Card project, with an estimate of Rs. 150,000 cr. In Sep-Oct 2000, the then Prime Minister had announced the same project under the banner of “National ID Card”, with an estimate of Rs. 6000 cr. On 26 June 2009, certain sections of the media (Like Siliconindia.com) had reported that this project had the estimate of Rs. 10,000 cr. In April 2009, the media had reported on the same project after talking to bidders of the ongoing PAN smartcard project, in which each card was supposed to cost Rs. 67. The latter card is similar to the UID Card except little inexpensive memory overheads. The centralized citizen’s date bank and related software for type-II query was also contained in the year 2000 project. So, the potential bidders, who have been in touch with the nodal agency – the NIC, had correctly reported the project estimate Rs. 10,000 cr.

The framework of year 2000 project had suggested a pilot in at least a city of size Delhi. The then PM Vajpayee had announced for pilot tests in 8 cities. Till date the government has not conducted any pilot in any sizable city or state. So, how could the government be confident of the estimate or the feasibility of the project? Small pilots aren’t enough; success of small-scale typing jobs was no proof of large-scale failure of 1994-95 voter-card project with more than 20% errors, as reported by the then acting CEC Krishnamurthy to the TOI in the autumn of 1996.

The year 2000 project was based upon a framework developed by the NIC, and the MHA (Home ministry), and its feasibility report was obtained from an Indian IT giant. Then the government had introduced the bill to the parliament. However, in Oct 2000 the Home ministry got convincing feedback that it was neither financially viable, nor legally nor technologically feasible, and so the bill, and the project were put into the cold storage. The NIC has been providing wisdom to the government on similar issues since the disastrous 1994-95 voter card project, in which Rs. 1000 cr was wasted. Several PSUs were contractors in many states then, and they also had more than 20% errors in 1994-95 voter card project.

So, the government should put the framework with its feasibility report on its website, so that the public can find out where and how the money is spent, and whether there are major loopholes in its feasibility.

]]> Comment on Nandan Nilekani to head UID Authority by Dr. Vijay Jha http://ss.emergic.org/2009/06/25/nandan-nilekani-to-head-uid-authority/#comment-19042 Dr. Vijay Jha Thu, 09 Jul 2009 08:43:19 +0000 http://ss.emergic.org/?p=1109#comment-19042 Omissions in the main article: 1. Issues of muslim's protest in a few states was not the main reason - it was a political reason. The main reason, as on record mentioned to the Home ministry in 2000 was that the framework of the project as developed and approved by the government was analysed to be financially non-viable, and technologically and politically not feasible. 2. The framework recommended to issue National ID Card in english, bypassing Indian languages because latter technology (OCR etc) was not developed at that time. They had forgotten the Rajbhasha Act of 1962, under which the state was bound to issue such document detailes in english as well as in another Indian language. Anyone could have filed a PIL to stop the project as it was. 3. The framework suggested that at least 4 year time was required to implement it under normal pace. They recommended around 15000 franchisee, each with 4 workstations and 4 data-capture operators. Assuming 1 year time for set-up and training, and assuming a adult population of 600 millions during 2001-2003, on average each franchisee will have to capture around 40000 ID Cards in 3 years, or 13333 cards per year, or 1111 cards per month with four operators. Assuming 23.3 working days per month as standard, we get less than 12 cards per operator per day. He will get Rs. 10 for data-capture of each card. So, each operator gets Rs. 120 per cards. The minimal wages for unskilled labour fixed by the government in general were Rs. 200 per day. So, operators were to get only 60% of the minimal wage - and this was not possible, and perhaps legally not viable in a government project. In the above we have to add franchisee's own salary, office rent, generators and its running cost in most parts of India, the cost of maintenance of equipment and office, the cost of investment by the franchisee, which the framework concludes to be 1.5 lacs, but a closer scrutiny of hidden costs shows it to be around Rs. 3 lacs, its interest and installments so that it is recovered within 3 years, and the profit at the franchisee level. Further, we have to account for the added cost of skilled labour for data-capture, especially photography and finger-print capture by sophisticated machines. If we add all this, his cost would increase to more than Rs. 35 per card. So, he was getting almost 3 times less than he should have got. 3. The irony is that in comparison to the 1994-95 Seshan's Voter Card project, the textual data to be entered in the National ID card was 3 times more, scanned items were more than 3 times more, photo was common in both, and a new item of fingerprint capturing was added. In 1994-95 the data capture cost was between Rs. 8-10 in most of the states. Assuming Delhi's cost of Rs. 8 for data capture, By 2001 price index it would be around Rs. 13. So, the national ID card had allocated 80% less to be paid for date capture, while it had more than 3 times more data to capture. It is clear that the quality of the job was to be poorer than the failed 1994-95 project. So, expect much more than 20% errors found during 1994-95 on average, across the country. 4. Finally, During autumn of 1996, the then acting CEC Krishnamurthy had reported to the TOI that there were more than 20% errors in 1994-95 Seshan's card project. Since an official document with such huge errors is a waste, the entire project was a grandiose failure. By that experiment, we find that the National ID card was to be a grandiose failure if implemented along the lines approved by the government. Since elections were due after 4 years, the government couldn't pursue it for a possible backlash. Legal hurdles due to Rajbhasha act was another issue. There were more technological issues. Conclusions: What were the reasons for such stupid errors in the government framework? The first reason was too much dependence upon the NIC, which has been doing very low-level government jobs, and lacks relevant expertise for deriving framework for such a huge project. T.N. Seshan too had implemented his 1994-95 Voter card project after borrowing wisdom from the NIC and a few public sector companies. Even "Typing job" is known to all, and yet they failed to analyze the cost of proof-reading, especially in such a large setup, where the law of large numbers plays vital role. They could have got opinion from the press on their own cost of proof-reading, or from small DTP-shop owners for 100% accurate jobs. Nothing of that sort was done. The terms of 1994-95 Voter-card tenders were incorrect - they asked for biddeers with turnover in several crores, but didn't ask whether this turnover was in the relavant area of data capture. The turnover must be subject-specific; a great potato-seller doesn't qualify for voter-card project. Furthet, the bidders subcontracted the data-entry part to small vendors, who also subcontracted it to small DTD-shop owners, small computer institutes, etc. So, the smallest unit which actually performed the job got pennys. They employed inexperienced operators at dismal wages, whence the quality. The operators were forced to correct errors by withholding their money - which was illegal. Under usual rates for typing, no operator or DTP shop owner will guarantee of 100% accuracy. Operator's speed comes down by 3-4 times if he is to verify each and every word. And yet there would be around 1% errors in words, or by the law of probability 8-9% errors in voter records. Proof-reading is supposed to mbe client's job in such setup. Or you pay for 100% accuracy, which will increase the cost by at least five time. Now the data-entry operators, during 1994-95, got 6 paise per citizen, and they were forced to correct the errors by withholding the payments. The big players got their money, in some cases 10%-20% deducted for poor show, and were on profit. If a bridge collapses, FIR is lodged against the supervising engineer as well as the contractor. In cases of election commission, no FIR has ever been lodged for such collapse of the entire project and waste of 1000 cr of public money. Democracy survives due to checks and balances by conscious citizens. In this cases no one bothered about details. Next thing is secrecy - why the government is not putting the framework and its feasibility report on its website, so that the public can find out where and how the money is to be spent, and whether it is feasible and viable? The national ID card project was worth Rs. 6000 cr. The UID card is little improved with PKI, but over 9 years the cost of cards have come down drastically. The similar card in the ongoing PAN-card project costs Rs. 67 each. Though salary have increased, the cost centre lies in technology - the card, and centralized printing, and the latter has come down. So, the overall total cost, including the cost due to 20% population growth, shouln't exceed 50%. Before the final government's announcement, some sections of the media, after talking to the potential bidders, estimated the cost to be Rs. 10000 cr. So, how did the government come to the figure of 150,000 cr? Finally, in Oct 2000, the then PM had announed that they will go for pilot in 8 cities. The framework of the project had recommended for a pilot in a city/state of at least the size of Delhi. Till date there has not been any pilot on such a size. Without such testing, how could the government be sure of financial viability and technical feeasibility? Omissions in the main article:
1. Issues of muslim’s protest in a few states was not the main reason – it was a political reason. The main reason, as on record mentioned to the Home ministry in 2000 was that the framework of the project as developed and approved by the government was analysed to be financially non-viable, and technologically and politically not feasible.
2. The framework recommended to issue National ID Card in english, bypassing Indian languages because latter technology (OCR etc) was not developed at that time. They had forgotten the Rajbhasha Act of 1962, under which the state was bound to issue such document detailes in english as well as in another Indian language. Anyone could have filed a PIL to stop the project as it was.
3. The framework suggested that at least 4 year time was required to implement it under normal pace. They recommended around 15000 franchisee, each with 4 workstations and 4 data-capture operators. Assuming 1 year time for set-up and training, and assuming a adult population of 600 millions during 2001-2003, on average each franchisee will have to capture around 40000 ID Cards in 3 years, or 13333 cards per year, or 1111 cards per month with four operators. Assuming 23.3 working days per month as standard, we get less than 12 cards per operator per day. He will get Rs. 10 for data-capture of each card. So, each operator gets Rs. 120 per cards.
The minimal wages for unskilled labour fixed by the government in general were Rs. 200 per day. So, operators were to get only 60% of the minimal wage – and this was not possible, and perhaps legally not viable in a government project. In the above we have to add franchisee’s own salary, office rent, generators and its running cost in most parts of India, the cost of maintenance of equipment and office, the cost of investment by the franchisee, which the framework concludes to be 1.5 lacs, but a closer scrutiny of hidden costs shows it to be around Rs. 3 lacs, its interest and installments so that it is recovered within 3 years, and the profit at the franchisee level. Further, we have to account for the added cost of skilled labour for data-capture, especially photography and finger-print capture by sophisticated machines. If we add all this, his cost would increase to more than Rs. 35 per card. So, he was getting almost 3 times less than he should have got.

3. The irony is that in comparison to the 1994-95 Seshan’s Voter Card project, the textual data to be entered in the National ID card was 3 times more, scanned items were more than 3 times more, photo was common in both, and a new item of fingerprint capturing was added. In 1994-95 the data capture cost was between Rs. 8-10 in most of the states. Assuming Delhi’s cost of Rs. 8 for data capture, By 2001 price index it would be around Rs. 13. So, the national ID card had allocated 80% less to be paid for date capture, while it had more than 3 times more data to capture. It is clear that the quality of the job was to be poorer than the failed 1994-95 project. So, expect much more than 20% errors found during 1994-95 on average, across the country.

4. Finally, During autumn of 1996, the then acting CEC Krishnamurthy had reported to the TOI that there were more than 20% errors in 1994-95 Seshan’s card project. Since an official document with such huge errors is a waste, the entire project was a grandiose failure.
By that experiment, we find that the National ID card was to be a grandiose failure if implemented along the lines approved by the government.
Since elections were due after 4 years, the government couldn’t pursue it for a possible backlash. Legal hurdles due to Rajbhasha act was another issue.
There were more technological issues.

Conclusions:
What were the reasons for such stupid errors in the government framework?
The first reason was too much dependence upon the NIC, which has been doing very low-level government jobs, and lacks relevant expertise for deriving framework for such a huge project. T.N. Seshan too had implemented his 1994-95 Voter card project after borrowing wisdom from the NIC and a few public sector companies. Even “Typing job” is known to all, and yet they failed to analyze the cost of proof-reading, especially in such a large setup, where the law of large numbers plays vital role. They could have got opinion from the press on their own cost of proof-reading, or from small DTP-shop owners for 100% accurate jobs. Nothing of that sort was done. The terms of 1994-95 Voter-card tenders were incorrect – they asked for biddeers with turnover in several crores, but didn’t ask whether this turnover was in the relavant area of data capture. The turnover must be subject-specific; a great potato-seller doesn’t qualify for voter-card project. Furthet, the bidders subcontracted the data-entry part to small vendors, who also subcontracted it to small DTD-shop owners, small computer institutes, etc. So, the smallest unit which actually performed the job got pennys. They employed inexperienced operators at dismal wages, whence the quality. The operators were forced to correct errors by withholding their money – which was illegal. Under usual rates for typing, no operator or DTP shop owner will guarantee of 100% accuracy. Operator’s speed comes down by 3-4 times if he is to verify each and every word. And yet there would be around 1% errors in words, or by the law of probability 8-9% errors in voter records. Proof-reading is supposed to mbe client’s job in such setup. Or you pay for 100% accuracy, which will increase the cost by at least five time.
Now the data-entry operators, during 1994-95, got 6 paise per citizen, and they were forced to correct the errors by withholding the payments.
The big players got their money, in some cases 10%-20% deducted for poor show, and were on profit.
If a bridge collapses, FIR is lodged against the supervising engineer as well as the contractor. In cases of election commission, no FIR has ever been lodged for such collapse of the entire project and waste of 1000 cr of public money.
Democracy survives due to checks and balances by conscious citizens. In this cases no one bothered about details.
Next thing is secrecy – why the government is not putting the framework and its feasibility report on its website, so that the public can find out where and how the money is to be spent, and whether it is feasible and viable?
The national ID card project was worth Rs. 6000 cr. The UID card is little improved with PKI, but over 9 years the cost of cards have come down drastically. The similar card in the ongoing PAN-card project costs Rs. 67 each. Though salary have increased, the cost centre lies in technology – the card, and centralized printing, and the latter has come down. So, the overall total cost, including the cost due to 20% population growth, shouln’t exceed 50%. Before the final government’s announcement, some sections of the media, after talking to the potential bidders, estimated the cost to be Rs. 10000 cr. So, how did the government come to the figure of 150,000 cr?
Finally, in Oct 2000, the then PM had announed that they will go for pilot in 8 cities. The framework of the project had recommended for a pilot in a city/state of at least the size of Delhi.
Till date there has not been any pilot on such a size. Without such testing, how could the government be sure of financial viability and technical feeasibility?

]]> Comment on Jai Prakash and Lok Satta Party – hope for better days ahead by Ravi kiran http://ss.emergic.org/2009/03/28/jai-prakash-and-lok-satta-party-hope-for-better-days-ahead/#comment-19041 Ravi kiran Thu, 09 Jul 2009 05:36:20 +0000 http://ss.emergic.org/?p=826#comment-19041 sir iam very happy that at least one person in our country wants to change our nation that is u iam always there with u to do any thing. sir iam very happy that at least one person in our country wants to change our nation that is u iam always there with u to do any thing.

]]> Comment on Jai Prakash and Lok Satta Party – hope for better days ahead by Ravi kiran http://ss.emergic.org/2009/03/28/jai-prakash-and-lok-satta-party-hope-for-better-days-ahead/#comment-19039 Ravi kiran Thu, 09 Jul 2009 05:32:55 +0000 http://ss.emergic.org/?p=826#comment-19039 Sir Namaste i like ur views and the way u think. its very Great to have such a Great person like u. sir one request i do have lots of things to do some iam doing i need ur appointment so i can discuss lots of things to u. Thank u very much urs Ravi kiran (Dont think what others have done for us always we have to think what we have done for others) Sir Namaste i like ur views and the way u think. its very Great to have such a Great person like u. sir one request i do have lots of things to do some iam doing i need ur appointment so i can discuss lots of things to u.

Thank u very much
urs
Ravi kiran
(Dont think what others have done for us always we have to think what we have done for others)

]]> Comment on Jai Prakash and Lok Satta Party – hope for better days ahead by Harish Nilopant http://ss.emergic.org/2009/03/28/jai-prakash-and-lok-satta-party-hope-for-better-days-ahead/#comment-19038 Harish Nilopant Wed, 08 Jul 2009 16:26:01 +0000 http://ss.emergic.org/?p=826#comment-19038 hello Jayaprakash Sir, Congratulation for your success, first time in my life i am proud to my vote. When you elected from kukatpally i really happy, and i want a give a flower bouquet to you but that time i cant dare. This is my first mail write to some politician in whole life. Today i saw your interview on zee telugu i really happy. You are giving such a nicely answer. paramatma (god) will help lot for your future success. Thanks You very hello Jayaprakash Sir,
Congratulation for your success, first time in my life i am proud to my vote.
When you elected from kukatpally i really happy, and i want a give a flower bouquet to you but that time i cant dare. This is my first mail write to some politician in whole life. Today i saw your interview on zee telugu i really happy. You are giving such a nicely answer. paramatma (god) will help lot for your future success.

Thanks You very

]]> Comment on Union budget size increases to Rs 1 Million Crores! by Manish Thaper http://ss.emergic.org/2009/07/06/union-budget-size-increases-to-rs-1-million-crores/#comment-19037 Manish Thaper Wed, 08 Jul 2009 14:02:02 +0000 http://ss.emergic.org/?p=1143#comment-19037 Respected Prof. Claims and allocation for poor and education had been made in past too. I feel this budget in no difference. Following is reproduced from Internet article and from your personal blog.Hope UID project {Headed by a renowned and well respected technocrat} may make a difference {a recent technological initiative in our country}. UID may aid in smooth transfer of money and government policies to the needful. In past crores of rupees that start journey from books preserved in National Capital had failed to reach the destined hands. I wish two years from now we the Indians plug those leaks. Help our current Prime Minister in achieving his dream of thinning the subsidy bill by restricting to the real needy and preventing it to reach the hands of greedy. Respected Prof.
Claims and allocation for poor and education had been made in past too. I feel this budget in no difference.

Following is reproduced from Internet article and from your personal blog.Hope UID project {Headed by a renowned and well respected technocrat} may make a difference {a recent technological initiative in our country}. UID may aid in smooth transfer of money and government policies to the needful. In past crores of rupees that start journey from books preserved in National Capital had failed to reach the destined hands.

I wish two years from now we the Indians plug those leaks. Help our current Prime Minister in achieving his dream of thinning the subsidy bill by restricting to the real needy and preventing it to reach the hands of greedy.

]]> Comment on Three Bangalore-based Institutes get NSF MIT pat by ssemergic http://ss.emergic.org/2009/06/10/three-bangalore-based-institutes-get-nsf-mit-pat/#comment-19034 ssemergic Mon, 06 Jul 2009 09:28:58 +0000 http://ss.emergic.org/?p=1067#comment-19034 Thanks Welcome to IIIT Thanks
Welcome to IIIT

]]> Comment on Three Bangalore-based Institutes get NSF MIT pat by allenki sridhar http://ss.emergic.org/2009/06/10/three-bangalore-based-institutes-get-nsf-mit-pat/#comment-19032 allenki sridhar Sun, 05 Jul 2009 15:53:37 +0000 http://ss.emergic.org/?p=1067#comment-19032 Hello sir, Hearty congratulations for achieving this great award .....and it is a very good news for me to hear this... because i have got admission recently in PGDSD program... from IIIT-B...so i had many doubts at the time of getting offer of my admission....but now....i am feeling very proud to be joining such a great institute where i can improve and acquire my IT skills and up to the industry level and going to interact with the great professors. Once again Congratulations...sir..and thanks for being me a student in your institute. Hello sir,

Hearty congratulations for achieving this great award …..and it is a very good news for me to hear this… because i have got admission recently in PGDSD program… from IIIT-B…so i had many doubts at the time of getting offer of my admission….but now….i am feeling very proud to be joining such a great institute where i can improve and acquire my IT skills and up to the industry level and going to interact with the great professors.

Once again Congratulations…sir..and thanks for being me a student in your institute.

]]> Comment on Nandan Nilekani to head UID Authority by A Hemmady http://ss.emergic.org/2009/06/25/nandan-nilekani-to-head-uid-authority/#comment-19031 A Hemmady Sun, 05 Jul 2009 04:39:36 +0000 http://ss.emergic.org/?p=1109#comment-19031 Thanks for your response. I am not opposed to the concept but it's not apparent or clear how it will be used and in what way it will contribute to the improve the lives of ordinary Indians. Thanks for your response. I am not opposed to the concept but it’s not apparent or clear how it will be used and in what way it will contribute to the improve the lives of ordinary Indians.

]]> Comment on Nandan Nilekani to head UID Authority by ssemergic http://ss.emergic.org/2009/06/25/nandan-nilekani-to-head-uid-authority/#comment-19030 ssemergic Sun, 05 Jul 2009 03:11:57 +0000 http://ss.emergic.org/?p=1109#comment-19030 Thanks for your interest and the trouble taken to comment UID is NOT about smart card; it is about an architecture to uniquely identify EVERY Indian. It is a basic requirement for any democracy to function Thanks for your interest and the trouble taken to comment
UID is NOT about smart card; it is about an architecture to uniquely identify EVERY Indian. It is a basic requirement for any democracy to function

]]>