The Digital India Campaign is another of Prime Minister Modi's pet projects. There is a lot of talk of laying optical fibre right up to each and every village in this country and providing country wide high speed internet access. Like much else, in this too there is more rhetoric than actual implementation, primarily because there isn't enough money to back up the loud mouthed announcements!!! For the past one and a half years or so we had been trying to get the internet to the Rani Kajal Jeevanshala (RKJ) school in Kakrana on the banks of the River Narmada in Alirajpur district of Madhya Pradesh without much success. Here is the detailed story of how we have now finally succeeded which tells a lot about what is wrong with this country!!
Let alone being connected with optic fibre, Kakrana does not have wireless connectivity also. The nearest mobile telephony tower is at a distance of 7 kms over hill and down dale. Thus, mobile phones don't work in Kakrana unless one climbs up on the hills. While most of the school buildings are in the valley below, there is one high hill about 30 meters in height within the campus of the RKJ. So whenever people have to communicate they climb up this hill with their mobiles. Smartphones can also receive a weak data signal of about 10 kilobites per second for checking emails at the most on this hill top. Since internet connectivity is so important in today's world and especially for a school for Adivasi children, initially a tent was set up on this hill to see if a laptop connected to a dongle wouldn't make possible some rudimentary emailing as shown below.
However, with a laptop the emailing did not take place smoothly mainly because the standard email providers like gmail and yahoo pack in a lot of advertisements with pictures which are too heavy for the very low speeds. When the emails are opened in smartphones then these advertisements are not there and so it is easier to access them at low speeds. So we began searching for alternatives. One was to go in for satellite internet but this had to be immediately rejected because it is very expensive. An attempt was made to get our research work on ecology in the school certified as a scientific activity so as to access free satellite internet through the ERNET facility of the government under the aegis of the Tata Institute of Fundamental Research. However, that did not materialise because NGOs are not eligible even if they are working under the aegis of a premier national research institute.
Looking around for solutions we came upon the wireless hop technology. This essentially meant using radio waves to transport data from the nearest point of good broadband optical fibre data connectivity in Kukshi some 35 kilometers away, straight as the crow flies, to Kakrana. The problem with this was that due to intervening hills there would have to be two repeaters on top of these hills between Kukshi and Kakrana. Setting up these repeaters and then ensuring their safety were both a cost and logistics issue. I went to see one such set up in Kodaicanal in the Nilgiri hills in Tamil Nadu to see how it works and the day I landed up it wasn't working because one of the repeaters had malfunctioned. This meant that someone would have to be sent to physically investigate the problem and set it right. Moreover, the software that has to be used for running this system is highly complicated and requires a fair level of computing skill which is not there in Kakrana. There are also some commercial operations in wireless hopping like Air Jaldi which is providing internet connectivity in the Himalayas in Himachal Pradesh and Uttarakhand but they said that they could make an economically viable foray in Kakrana only if there was a substantial customer base prepared to pay for the service. So that option also fizzled out as the Bhils of the area around the school living at subsistence levels were unlikely to pay for internet service!! Indeed Air Jaldi has been stymied in its attempts to expand its base in the Himalayas precisely because it isn't getting enough paying customers.
Just when we thought that there was no solution I saw a Facebook post by Arjun Venkatraman of the Mojolab Foundation about a pilot he was setting up in a rural area of Karnataka with the wireless mesh technology. This involved setting up a local area network (LAN) with wireless routers with one of the routers having internet connectivity which he calls the community owned wireless mesh or COWMesh. This can work only if there are a number of enthusiastic participants in the community who are interested in keeping the mesh active and running. Arjun combines the COWMesh with Datamuling. Since in remote locations internet speeds are abysmally low it is not possible to do much more than emailing and that too through a client that edits out all the advertising that has heavy data. So educational content has to be downloaded at locations where data is cheap like in the Metros, put into hard disks and then transported by rail or road to remote locations where they can they be fed into the COWMesh for sharing on the LAN. On our invitation Arjun came down to Kakrana in April with a demonstration set of wireless routers and a raspberry pi which is a rudimentary computer which acts as the e-mail server when a dongle is connected to it. The e-mail server and one router was put on top of the hill and the other router was put at the bottom in the office and it gave the WiFi signals which then allowed other devices to access the internet. The first ever email from the office of RKJ was sent and it even had attached photographs. So finally a solution had been found and it was decided that Arjun would prepare a system for RKJ and set it up. Arjun did not charge for his expertise as he is funded by the Ashoka Foundation as a fellow to use innovative technology to spread internet and mobile connectivity in rural areas and he charged only for his travel and the cost of equipment which totally came to about Rs 40,000 which is a steal.
However, there were more problems to be overcome. Like most other high tech electronic equipment there is constant innovation in wireless routers also. Therefore, when the new routers arrived Arjun found that their hardware and software had both been changed in such a way that it was not possible to put into them the open source software that Arjun and some other wireless connectivity experts had developed for the COWMesh. He tried to hack in and change the configuration but this blocked the routers altogether as they had been password protected. Basically, the router manufacturer was trying to ensure that only their proprietary software would be used and open source configurations could not be superimposed. Then issued a lengthy email battle trying to get the manufacturer to replace the routers. It appears that the password lock had been so designed that even the manufacturer could not restore the software onto the routers once the router was locked!! Anyway, after much argumentation, the manufacturer replaced the routers with a new set. Arjun then prepared a COWMesh software that would sit over the routers in the raspberry pi.
Last week he came down to Kakrana with the modified system and then began the process of setting it up. LAN cables were required to connect the various parts of the system. These cables cost Rs 150 per metre mainly because of the expertise of crimping the connectors at the end which has to be done properly. However, Arjun had other ideas as he bought the cable at Rs 7 per metre and the connectors at Rs 3 each separately and a crimping tool. He taught Gulab and Dhani, the two operations and maintenance people at RKJ, to crimp the connectors on to the cable as shown below and so we had LAN cables at Rs 15 per metre!!
The routers had to be set up on pipes and covered with buckets to ensure that they were not affected by rain and sun. This fabrication work too was done by the duo of Gulab and Dhani as shown below using PVC water pipes and connections.
Finally one router along with the server had to be put on top of the water supply tank that had been built on top of the hill to get the mobile signal as shown below. The server construction involved some more local innovation as it was housed in a steel box which is sturdy enough to withstand both the sun and the rain. The lid is shown open in the picture to show the raspberry pie and dongle but it is actually kept closed.
The power supply is provided through an UPS and inverter. However, since this can provide a back up of about 8 hours only and sometimes the grid power outage in Kakrana is for much more than that, the next step is to put in a hybrid grid cum solar power unit that can power the whole IT set up in Kakrana involving the server, routers, laptop and printer. The COWmesh is now up and running and for the first time many teachers and students in Kakrana have sent and received their first emails. Once it was clear in April that sooner or later the COWmesh would be up in Kakrana, over the past two months a huge amount of educational content has been downloaded in Indore and transported by hard disk to Kakrana. So now regularly this content is shared on the COWmesh and the children get to see educational videos through a projector every evening. Since, the assembly of the COWmesh and its running have now been perfected by the staff of RKJ, they can easily set up other such meshes anywhere in the country. The Mojolab Foundation will provide the software and the hardware and RKJ will do the installation and set up. In future the COWMesh can be extended to the villagers in Kakrana also and they too can watch the educational content on their TVs if they are ready to pay the nominal price of running the system.
While at the community level the solution has been found and that too extremely cheaply and in the process the people have been technologically empowered, the problem of low speed of wireless internet connectivity in rural areas remains. It would be educative to go into the reason why this is so. The main reason is the adverse economics of mobile towers. It takes Rs 40 Lakhs to set up a mobile tower. In rural areas the grid electricity supply is unreliable and so the towers have to be powered by diesel generators. If good data speeds are to be provided then the towers have to be powered much more than just for voice connectivity and this will require a higher expenditure on diesel. The demand in rural areas is mostly for voice, that is, people are prepared to pay for voice and not for data. Therefore, to earn profits after having made huge investments in buying spectrum, the telecom companies first economise on building towers and then economise on the diesel for running those towers by under powering them and that is why in Kakrana the average data speed is only about 10 kbps. Clearly, the telecom companies including the Government owned Bharat Sanchar Nigam Limited are not going to improve wireless data connectivity unless they are given a huge subsidy to do so. Similarly the Bharat Broadband Nigam Limited which is tasked with providing optic fibre connectivity to rural areas is also unlikely to do so unless there is a huge subsidy for the capital and operational costs as poor people in rural areas are unlikely to bear this cost. The Telecom Regulatory Authority of India has initiated a public consultation on a new Public Wifi policy for spreading internet use but it does not adequately define public Wifi to include community meshes like the one set up in RKJ. Neither is there any support in terms of reducing duties on imported equipment necessary for setting up such networks. Presently community networks for sharing of content or accessing the internet can be set up without any regulatory permissions but if such permissions become mandatory in future then it will only make the task of digitising India that much more difficult as it is inconceivable that communities will be able to go through the application process for such permissions without costly mediation by legal experts.
There is still a problem that cannot be immediately solved because of a regulatory hurdle. Voice telephony cannot be transferred in real time from the top of the hill to the school below through a gateway even though it is technologically possible because of a legal ban. So what is being done as a next step in Kakrana is that a voice mail box system is being installed which will allow exchange of voice messages through a gateway with a time lag but won't allow real time conversations.
All this just shows that the Government in this country does not back up its rhetoric with regard to digitising India with adequate policy and financial support to really reach the internet to the poor. After all, after having earned huge sums of money by selling spectrum to the Telcos, it is surely obliged to spend some of that money to digitise India. But that is a distant dream.
Let alone being connected with optic fibre, Kakrana does not have wireless connectivity also. The nearest mobile telephony tower is at a distance of 7 kms over hill and down dale. Thus, mobile phones don't work in Kakrana unless one climbs up on the hills. While most of the school buildings are in the valley below, there is one high hill about 30 meters in height within the campus of the RKJ. So whenever people have to communicate they climb up this hill with their mobiles. Smartphones can also receive a weak data signal of about 10 kilobites per second for checking emails at the most on this hill top. Since internet connectivity is so important in today's world and especially for a school for Adivasi children, initially a tent was set up on this hill to see if a laptop connected to a dongle wouldn't make possible some rudimentary emailing as shown below.
Looking around for solutions we came upon the wireless hop technology. This essentially meant using radio waves to transport data from the nearest point of good broadband optical fibre data connectivity in Kukshi some 35 kilometers away, straight as the crow flies, to Kakrana. The problem with this was that due to intervening hills there would have to be two repeaters on top of these hills between Kukshi and Kakrana. Setting up these repeaters and then ensuring their safety were both a cost and logistics issue. I went to see one such set up in Kodaicanal in the Nilgiri hills in Tamil Nadu to see how it works and the day I landed up it wasn't working because one of the repeaters had malfunctioned. This meant that someone would have to be sent to physically investigate the problem and set it right. Moreover, the software that has to be used for running this system is highly complicated and requires a fair level of computing skill which is not there in Kakrana. There are also some commercial operations in wireless hopping like Air Jaldi which is providing internet connectivity in the Himalayas in Himachal Pradesh and Uttarakhand but they said that they could make an economically viable foray in Kakrana only if there was a substantial customer base prepared to pay for the service. So that option also fizzled out as the Bhils of the area around the school living at subsistence levels were unlikely to pay for internet service!! Indeed Air Jaldi has been stymied in its attempts to expand its base in the Himalayas precisely because it isn't getting enough paying customers.
Just when we thought that there was no solution I saw a Facebook post by Arjun Venkatraman of the Mojolab Foundation about a pilot he was setting up in a rural area of Karnataka with the wireless mesh technology. This involved setting up a local area network (LAN) with wireless routers with one of the routers having internet connectivity which he calls the community owned wireless mesh or COWMesh. This can work only if there are a number of enthusiastic participants in the community who are interested in keeping the mesh active and running. Arjun combines the COWMesh with Datamuling. Since in remote locations internet speeds are abysmally low it is not possible to do much more than emailing and that too through a client that edits out all the advertising that has heavy data. So educational content has to be downloaded at locations where data is cheap like in the Metros, put into hard disks and then transported by rail or road to remote locations where they can they be fed into the COWMesh for sharing on the LAN. On our invitation Arjun came down to Kakrana in April with a demonstration set of wireless routers and a raspberry pi which is a rudimentary computer which acts as the e-mail server when a dongle is connected to it. The e-mail server and one router was put on top of the hill and the other router was put at the bottom in the office and it gave the WiFi signals which then allowed other devices to access the internet. The first ever email from the office of RKJ was sent and it even had attached photographs. So finally a solution had been found and it was decided that Arjun would prepare a system for RKJ and set it up. Arjun did not charge for his expertise as he is funded by the Ashoka Foundation as a fellow to use innovative technology to spread internet and mobile connectivity in rural areas and he charged only for his travel and the cost of equipment which totally came to about Rs 40,000 which is a steal.
However, there were more problems to be overcome. Like most other high tech electronic equipment there is constant innovation in wireless routers also. Therefore, when the new routers arrived Arjun found that their hardware and software had both been changed in such a way that it was not possible to put into them the open source software that Arjun and some other wireless connectivity experts had developed for the COWMesh. He tried to hack in and change the configuration but this blocked the routers altogether as they had been password protected. Basically, the router manufacturer was trying to ensure that only their proprietary software would be used and open source configurations could not be superimposed. Then issued a lengthy email battle trying to get the manufacturer to replace the routers. It appears that the password lock had been so designed that even the manufacturer could not restore the software onto the routers once the router was locked!! Anyway, after much argumentation, the manufacturer replaced the routers with a new set. Arjun then prepared a COWMesh software that would sit over the routers in the raspberry pi.
Last week he came down to Kakrana with the modified system and then began the process of setting it up. LAN cables were required to connect the various parts of the system. These cables cost Rs 150 per metre mainly because of the expertise of crimping the connectors at the end which has to be done properly. However, Arjun had other ideas as he bought the cable at Rs 7 per metre and the connectors at Rs 3 each separately and a crimping tool. He taught Gulab and Dhani, the two operations and maintenance people at RKJ, to crimp the connectors on to the cable as shown below and so we had LAN cables at Rs 15 per metre!!
The routers had to be set up on pipes and covered with buckets to ensure that they were not affected by rain and sun. This fabrication work too was done by the duo of Gulab and Dhani as shown below using PVC water pipes and connections.
While at the community level the solution has been found and that too extremely cheaply and in the process the people have been technologically empowered, the problem of low speed of wireless internet connectivity in rural areas remains. It would be educative to go into the reason why this is so. The main reason is the adverse economics of mobile towers. It takes Rs 40 Lakhs to set up a mobile tower. In rural areas the grid electricity supply is unreliable and so the towers have to be powered by diesel generators. If good data speeds are to be provided then the towers have to be powered much more than just for voice connectivity and this will require a higher expenditure on diesel. The demand in rural areas is mostly for voice, that is, people are prepared to pay for voice and not for data. Therefore, to earn profits after having made huge investments in buying spectrum, the telecom companies first economise on building towers and then economise on the diesel for running those towers by under powering them and that is why in Kakrana the average data speed is only about 10 kbps. Clearly, the telecom companies including the Government owned Bharat Sanchar Nigam Limited are not going to improve wireless data connectivity unless they are given a huge subsidy to do so. Similarly the Bharat Broadband Nigam Limited which is tasked with providing optic fibre connectivity to rural areas is also unlikely to do so unless there is a huge subsidy for the capital and operational costs as poor people in rural areas are unlikely to bear this cost. The Telecom Regulatory Authority of India has initiated a public consultation on a new Public Wifi policy for spreading internet use but it does not adequately define public Wifi to include community meshes like the one set up in RKJ. Neither is there any support in terms of reducing duties on imported equipment necessary for setting up such networks. Presently community networks for sharing of content or accessing the internet can be set up without any regulatory permissions but if such permissions become mandatory in future then it will only make the task of digitising India that much more difficult as it is inconceivable that communities will be able to go through the application process for such permissions without costly mediation by legal experts.
There is still a problem that cannot be immediately solved because of a regulatory hurdle. Voice telephony cannot be transferred in real time from the top of the hill to the school below through a gateway even though it is technologically possible because of a legal ban. So what is being done as a next step in Kakrana is that a voice mail box system is being installed which will allow exchange of voice messages through a gateway with a time lag but won't allow real time conversations.
All this just shows that the Government in this country does not back up its rhetoric with regard to digitising India with adequate policy and financial support to really reach the internet to the poor. After all, after having earned huge sums of money by selling spectrum to the Telcos, it is surely obliged to spend some of that money to digitise India. But that is a distant dream.
No comments:
Post a Comment