Disproportionate costs and operational irritants force the government to call off the Media Lab Asia programme, a collaborative venture with the U.S.-based MIT that was supposed to bring in cutting-edge technology and help plug the digital divide between the country's urban and rural areas.
THE brainchild of former Minister for Communications and Information Technology Pramod Mahajan, Media Lab Asia (MLA) was launched in June 2001 with a lot of hype about India receiving cutting edge information and communication technologies (ICT) to solve problems at the grassroots level. But on May 11, the current Information Technology Minister, Arun Shourie, made public the government's decision to terminate its 18-month, high-profile collaboration with the Massachusetts Institute of Technology (MIT). Shourie's decision overrode the recommendation made by the MLA Evaluation Committee in December 2002 to extend the programme for 10 years.
Based on the testimonies of participating research teams from the five Indian Institutes of Technology (IITs) indicating that the value addition and inputs from the MIT to the projects being undertaken under the umbrella of MLA were minimal, and certainly not commensurate with the hefty sums paid to the MIT, the Minister sought to end the collaboration agreement. Shourie said: "The expectations on which MLA was based were not being realised. The tie-up was expected to bring in private investment, which did not materialise. The contributions by MIT were not in evidence and MIT had not been put to any specific deliverables at the end of the year" (see interview). "Basically, the collaboration agreement was not renewed after the expiry of the exploratory phase on March 31," Shourie clarified.
However, the MIT's version of the events is quite the opposite. On May 9, Professor Nicholas Negroponte, chief of the MIT Media Lab (MIT-ML), on which MLA was sought to be modelled, had pre-empted Shourie by telling the U.S. media that the MIT had decided to pull out of the agreement because the new Minister was not interested in uplifting India's rural poor. "He wants a very directed, project-oriented research and that is not the way we work at Media Lab," Negroponte was quoted as saying. He had further stated that he would set the record straight soon. The trigger for the eventual calling off of the project seems to have been the demand made by the MIT for an additional payment of $5 million or $1 million a year over 10 years for the use of the brand name, Media Lab. Shourie rejected the demand and began a review process with IIT research groups. Curiously, the demand for the payment over and above the agreed amount of $1.7 million for the period beyond the exploratory phase had been made even before the government approved the second phase of the project.
The turn of events is likely to give rise to the impression that Shourie is undoing what his predecessor had established. The collaborative venture was touted as one of Mahajan's crowning achievements. The arrangement would bring the MIT's brand equity to ICT-based projects meant to benefit India's rural masses and bridge the digital divide between urban and rural areas, it had been claimed.
The MIT-ML was founded in 1985 by Negroponte and Jerome Wiesner, former president of the MIT and science adviser to President John F. Kennedy. "Inventing the future," is the vision of Negroponte and the objective of the MIT-ML's multi-disciplinary research, which is driven by emerging communication and information technologies. MLA is the second Media Lab outside the U.S., the first one being Media Lab Europe (MLE), set up in Dublin, Ireland, in July 2000. The basic idea of MLA is to leverage through collaboration the MIT-ML's "entrepreneurial" model and its "innovative" consortia approach to research. It aims to adapt the research methodology to inter-disciplinary research projects, particularly those that would have an impact on India's rural population. While the MIT-ML is today credited with the invention of path-breaking information technologies such as MP3 music and e-ink, it is also known for its extravagant style of doing research. Its $40-million annual budget is funded to the tune of 95 per cent by corporate sponsors whose number in 2000 was 176 including ICT majors such as Microsoft, Intel, Motorola, British Telecom and Nokia. The Lab has 30 research teams working on over 250 projects.
THE idea of replicating the concept in India did not grow out of any felt need of the Indian scientific community; it was the realisation of a powerful politician's desire. During Mahajan's visit to the U.S. in June 2000 as part of an information technology (IT) delegation, he was greatly impressed by the Lab and by the presentation made by its research chief, Alex Pentland. Having just established MLE, the MIT-ML was exploring the possibility of setting up similar outposts in Asia and might have proposed one in India. Citing the prospect of bringing the latest in IT, coupled with the argument that if India missed the chance it would be grabbed by countries such as China and Singapore, the idea was sold easily to the government. Once the political decision had been made, and given the standing that Mahajan enjoyed within the government, nothing could have prevented MLA from coming up. At one of the preparatory meetings, Mahajan was reported to have said: "If there is a Taj Mahal there, why can't we have it here?"
A committee under V. Rajaraman, Professor of Computer Science at the Indian Institute of Science (IISc), Bangalore, was appointed to look into the feasibility of establishing MLA in collaboration with the MIT-ML. Pertinently, the committee asked what the Indian side was seeking from the MIT and what the latter had to offer in the Indian context. It observed that private sponsorship, which was the key to the model in the Indian context, might not be easily forthcoming, given the private sector's lack of interest in funding research and development (R&D) with a long-term perspective.
Even though these issues were never satisfactorily answered, the project was pushed through. The two issues raised by the Rajaraman Committee lie at the core of the collapse of the partnership and the present fate of MLA.
In August 2000, on the Minister's directive and as a follow-up to the Rajaraman Committee report, a senior official from the Department of Information Technology (DIT) visited the MIT-ML to understand fully its business model, its `shared IPR (Intellectual Property Rights)' concept according to which all the private sponsors have equal IPRs over the technologies developed by the MIT-ML and, most important, how the Lab functioned. In November 2000, even as the DIT was evaluating the proposal, an MIT team led by Negroponte, which included a legal adviser, descended on Delhi to initial the agreement with the government. Clearly, there was a parallel channel of communication between the Minister and the MIT-ML. It was decided that the collaboration will be for a 10-year period with an initial, exploratory phase of one year when the project would be funded entirely by the government. The decision to continue the partnership could be taken after assessing the performance in this phase. The investment involved in Negroponte's original suggestion, $30 million in the first year, was scaled down to Rs.65 crores (about $14 million). Also, Negroponte was told that corporate sponsorship might not be forthcoming as envisaged.
However, representatives of the MIT assured the government that corporate sponsorship would come as the programme progressed and recommended a total investment of $1 billion for the 10-year period, with a commitment of 20 per cent of the amount from the government. Apparently they argued that for a vast country like India the scale of operations should be big and that the extent of the potential reach of technologies generated in terms of volume would automatically bring in the remaining 80 per cent of the funds from non-governmental sources. How this huge figure, which adds up to about three times the investment at the MIT-ML, was arrived at is not clear. Strangely enough, Mahajan fell for the argument and the arbitrarily quoted figure. The interventions made by DIT officials in the course of the discussions were apparently of no avail.
Soon, a Joint Task Force (JTF) of the government and the MIT was constituted to draw up an action plan to establish MLA and identify its potential research areas. In the JTF Report of February 2001, DIT officials identified the risk factors inherent in the venture. The report pointed out that the complexity and sheer scale of the project could be "its own enemy with too many players pulling in different directions". Further, the report stated that the MIT-ML did not have a background of executing field-level projects on a scale envisaged by MLA. "MIT-ML may be overstretched in terms of its resources. The Ministry may also find it difficult to handle projects of the kind envisaged here," the report said. In hindsight, the incorporation of a trial phase of one year was a wise move because some of the potential problems identified by the JTF seem to have surfaced in the first phase itself.
Clearly, the programme enjoyed political backing at the highest level. Notwithstanding its scale, the project was one of the quickest to be cleared by the Cabinet. It was approved by the Cabinet on May 31, 2001, barely three weeks after it had been put up. The agreement was signed on June 24, 2001, and MLA was registered as an Indian company under Section 25 of the Companies Act on September 20, 2001. As per the Research and Collaboration Agreement (RCA), which was signed on September 21, the MIT was to be paid $1.7 million (Rs.9.5 crores) net of taxes as compensation for expenses such as travel, communications, salaries, indirect costs and fellowship expenditures that were incurred. The agreement envisaged an investment of Rs.5,127.50 crores, about 83 per cent (Rs.4,257.50 crores) of which would be raised from sponsors in industry and other non-governmental sources, including international donor agencies. The remaining Rs.870 crores would be in the form of governmental grant-in-aid. The programme envisaged creating a federal network consisting of laboratories, non-governmental agencies (NGOs), academics and other organisations and end users. Mahajan hails from Maharashtra, and it is significant that MLA had its headquarters in Mumbai.
In the initial phase, five IITs agreed to participate and act as MLA centres, in Mumbai, Delhi, Kanpur, Kharagpur and Chennai respectively. It is a moot question whether the collaboration would have been scrapped if the Minister had not changed or if the chance investigation by Shourie had not revealed the fundamental problems in the management of MLA, especially after the evaluation committee's report saw "good progress" and "momentum" in the first year. The committee recommended that the collaboration be extended for the next 10 years in its full scope, albeit with a scaled-down and "realistic" investment target of Rs.2,400 crores, and a pan-Asian character.
In the light of views expressed by the IIT researchers and in view of the fact that private sponsorship had not been realised at all, it is far from clear how the committee, with an IIT director on board, came to such a conclusion. It was clear from the beginning that it would be difficult to attract private investment: as was pointed out by the Rajaraman Committee report and supported by the DIT's own internal assessment. It is easy to see that what might work in the MIT-ML's context is unlikely to work in the case of MLA. Raising about $30 million from 180 multinational ICT firms is not a big task. But in the Indian context, garnering a sum of Rs.100 crores to 200 crores from private sources means the involvement of at least 100 domestic firms, each with Rs.1 crore to invest and with no immediate "big bang for the buck". Clearly, this was totally unrealistic, especially in the wake of the IT meltdown and at a time when even the MIT-ML was finding it hard to raise money from its trusted sponsors. Indian companies would rather invest this sum in 10 employees. Except for the Rs.1.5 crores that came from Tata Consultancy Services (TCS), no contribution was forthcoming, even from companies such as Infosys, whose chairman was on the board of directors of MLA. For multinational companies (MNCs), given the kind of projects, it was not an attractive proposition.
Chief among the operational problems was the conflict inherent in the styles of functioning of the MIT-ML and Indian research groups, with overheads, expenses not related to research, and salaries far exceeding the scales of operation that the research culture in the Indian government sector is used to. Also, the MIT-ML's culture of counter-intuitive, freewheeling and risk-taking research was sought to be imposed on MLA's projects, with no deliverables defined that could be evaluated against at the end of the year. Indeed, no monitoring system to assess the progress of the programmes was put in place. Even though the exploratory phase was funded entirely by the government to the tune of Rs.65 crores, it seems that the MIT-ML was seeking to control the entire operation even when it did not make any significant contribution to the projects in terms of ideas, hardware or technology. Interestingly, MLA's chief executive officer (CEO) Bimal Sareen, a marketing executive from Compaq, was not an appointee of the government. He was the MIT's choice and had been identified by a head-hunter firm Heidrick & Struggles. And the MIT had been paid Rs.9.5 crores for its "services". As a senior DIT official put it, it amounted to "authority without responsibility".
The resignation of S. Ramakrishnan, the DIT's representative on the management committee, in April 2002 was evidence of the undermining of the DIT's control over MLA's operations. The lack of a leadership committed to delivering results, one that would be in charge and in total control of MLA's operations, was perhaps a glaring lacuna in the management structure. The flaw lay in expecting a non-technical CEO to give directions to a federally instituted research network. The appointment of the CEO was made only in June 2002.
Was the programme ill-conceived? "No," says Krithi Ramamritham, the MLA project coordinator at the IIT-Powai hub. "It was well-conceived from the point of view of research. The part that I think was not fleshed out fully was the one relating to the deliverables from the MIT end. They were supposed to be consultants but we, as researchers, did not know what we can ask them for. The interaction was in some sense one-way. They came here, gave talks, listened to what we were doing these projects had been arrived at locally and had been going on before and gave a few comments. The problem, I think, was that what was expected from them was not clear to any of the people concerned." If this was the case, it is pertinent to ask why the IITs agreed to collaborate, especially if the projects had been conceived independent of MLA and had been going on in any case.
Even though nothing of significance came from the MIT, MoUs (memorandums of understanding) were entered into with MLA; according to informed sources, the primary reason appears to be the attraction of big money. From the perspective of the IITs, of the Rs.65 crores, leaving aside about Rs.15 crores for MLA's operations and the MIT's expenses, there was about Rs.10 crores for the taking by each IIT in one year for R&D as well as downstream activities, something that was unusual but certainly welcome. Also, there was potential for collaboration even if undefined with the MIT and the IITs. But when the quantum of money that eventually came was not up to their expectations and they realised that disproportionate sums were being spent on the MIT and other non-research related activities, the IITs seem to have got irked.
Consider the following: As of October 2002, of a total of Rs.22.86 crores that was spent, payments to the MIT alone accounted for 41.54 per cent, while the amounts spent on all the IITs and downstream research activities, including field demonstrations, accounted for just 34.46 per cent. Expenses relating to activities other than research (Rs.5.48 crores) accounted for 24 per cent. Among the IITs, the highest amount, Rs.1.62 crores, was given to IIT-Kanpur and the lowest amount, Rs.37 lakhs, to IIT-Kharagpur. As of March 31, a total of Rs.33 crores had been spent, and the proportion would have remained roughly the same.
Ashok Jhunjhunwala, the project coordinator at IIT-Chennai, says: "Actually, MIT Media Labs contributed nothing at the technical level and was trying to play big brother, telling us what to do. On top of it, we found that it was creating huge hype. All this would have been still tolerated until we found out that we were paying MIT cash, in addition to taking care of its expensive ways. The team that it created to manage Media Lab Asia was also costing a huge amount." Curiously enough, none of these points seems to have been brought to the notice of the Technical Advisory Group (TAG) whose co-chair, incidentally, was the Director of IIT-Kanpur, Sanjay Dhande.
The argument being advanced is that the researchers were unaware of the chemistry of the operations in MLA. The teams were never specifically asked for inputs on how the collaboration was going on, nor were they asked whether they benefited from interactions with the MIT team. According to members of the teams, had they been asked earlier, they would have taken time out from their research work and given inputs. But why Dhande himself did not raise the issue at any level remains unclear. Dhande was not available for comment on the issue.
Giving an overall assessment, Ramamritham says: "Too many things were happening at the same time. Research was beginning, taking shape and getting crystallised. Before that happened, sponsors were being sought. And sponsors asked `what do we get for our money?' There was far too much pressure on the management to bring in sponsorship rather than focus on research. And one year is too short a time to learn from mistakes and do a mid-course correction."
SO what is the future of MLA? Based on inputs from the research groups, the Minister has drawn up a proposal with a remoulded structure to create a scheme that will retain the MLA model of collaborative research. Ideation arrived at by the brainstorming of a consortium of researchers, stake-holders including non-governmental parties, and end-users is a new concept for Indian researchers and may be borrowed from the MIT-ML's concept, feels Ramamritham. "A lot of mistakes were made. Change was badly needed. Of course, there will be some confusion and the projects will get a little hurt. We need not swing the other way round, but rework ways of doing this. In the end, the MLA concept will emerge as a winner," feels Jhunjhunwala.
There are critics of the very model on which MLA was established. Sachin Maheshwari, Professor of Computer Science at IIT-Delhi, says: "The proposal to continue the MLA model by replacing the MIT by the IITs also will not work. MLA laboratories on-campus is attractive because it is easy money. There will not be and there cannot be any commitment from the IITs to ensure that MLA succeeds because its success or failure does not impact the IITs at all. If MLA is to continue, its laboratories should be in institutions other than the IITs. There are a number of institutions that have a stake in such projects and where MLA in the form of a laboratory will provide that additional support and inputs to the work being done, which at the present time is missing."
In the new proposal awaiting Cabinet clearance, a funding of Rs.900 crores, of which Rs.250 crores will be from private sources, has been envisaged for the 10th and 11th Plan periods, each of which will receive Rs.262 crores and Rs.35 crores respectively, including Rs.32 crores left over from MLA's 2002-03 budget. This funding will be outside the nominal DIT budgetary allocations. The Finance Ministry is not too happy with the ratio of government to private funding and the Planning Commision does not want to create a new entity but would like it to operate as a scheme under the DIT. How the Cabinet decides on the matter remains to be seen.