Public WiFi in Delhi: Some Suggestions

21 March 2015

Recently the government in Delhi formed a task force for policy suggestions for the free public Wifi promise they gave during the Delhi elections. This is a welcome step, especially since they have chosen some good people at the helm - Nikhil Pahwa from MediaNama, Mahesh Murthy from Pinstorm, and Pranesh Prakash from the Center of Internet and Society.

One thing to note is that they are just an advisory group - their recommendations will ultimately be upto the government to accept and implement. I don't expect the government to fully accept each and every one of the recommendations, so the end result will hinge on which public policy recommendations the government accepts, and what the level of implementation (and further upkeep and maintenance) is.

Mahesh posted some questions and sought feedback from the community on those. Here, I'll try to do my part and add my two cents on those questions.

1. How do you define public wifi? Kindly suggest case studies on public wifi from other parts of the world.

You can get really technical here, and it's a somewhat complicated task to even define WiFi, but for the purposes of this policy work, I think it could be simply 'publicly available wireless internet access'.

There are a number of case studies around of the world where cities have either implemented public WiFi, either in a small area, or a rather large area. Some case studies can be found in this white paper from the Wireless Broadband Alliance.

The city of Langley, British Columbia in Canada also implemented public WiFi called 'B-NET' (in a limited area) at relatively low cost by partnering with a Hungary based Hotspot provider called HotSpotSystem. The city of Vancouver has a proposal to have 'V-Poles' (akin to street lights, but with added infrastructure for smart services) throughout their city. New York City had announced it's attempt to implement public WiFi - initially starting in a limited area (including subway stations) and using (at least in-part) private donations. You can know more about it by going to the LinkNYC site and reading this article detailing how LinkNYC works.

However, I personally recommend looking at Seoul, along with New York City, as a model of public WiFi access for Delhi in terms of implementation. The city of Seoul has one of the best WiFi coverages in the world, and is fast and consistent. Here is one more case study on Seoul as a smart city and here too. The Model for NYC is an advertisement supported model, which could also prove sustainable and cost effective.

In terms of services on top of WiFi that the government can provide, as well as the general mindset of the government in relation with the internet, I would suggest taking a look at what Estonia has been doing with the Internet. Also, please look at Singapore's Environment Minister Mr. Vivian Balakrishnan and how Singapore approaches open data from the government's perspective.

2. What should be the ideal level of access to public wifi? Can it be made available right upto your doorstep? If yes, what are the feasibility and technological challenges? Or should it be made available at well-defined public spaces in every neighbourhood?

I think it will be unrealistic to make it available to everyone's doorstep given a range of factors like the size of Delhi, the amount of buildings, concrete walls and basements (which often hinder signals). We could do it in a second or third phase of the rollout, but initially, I think it makes sense to roll it out in well-defined public areas.

This would make implementation much easier, and also maintenance much easier as if there is a problem, it would be easier to identify the source and fix it. As it is, most people don't require wifi just as they head out their doorstep - they need it at bus stops, train stations, metro stations (including underground subway stations especially as network coverage underground is bad), airports, shopping malls, parks, public libraries, museums, historical monuments, major public market areas etc. These are enough areas already, and implementation and maintenance of the service in these areas alone will require major effort and cost.

Once we implement these areas, and get further lessons and learning from this phase, we could think of rolling it out further.

3. What is the infrastructure and technology required to create the public wifi network?

This will depend on the kind of approach which the government goes for in the end. In any case though, the addition of hotspots would be required all over the city. This would involve installing a large number of WiFi routers and repeaters in the city.

We could look into re-using existing cell-phone towers to place hotspots wherever possible. An NYC-like approach would also require setting up kiosks, which initially might require significant investment, but in the long run might be cost neutral as it would support advertising.

4. What should be the essential services that a government should aim at providing through public wifi?

I think the number one use case from the government's perspective needs to be dissemination of information in case of emergencies and disasters (Medical outbreaks, terrorist threats, flood or cyclone warnings, etc.). A service to request emergency medical, fire-fighting and police help online would also be thought about. (Geolocation services on the web can often provide a much better estimate of location, which would help these services in reaching in time better. Also, it would help tourists in distress who might not know the area they are in.)

Another lower priority, but still important category would include real-time traffic information (especially as Delhi is a VIP-heavy area, with plenty of roads being blocked often due to International, Central, and state dignitaries), public holiday announcements, weather forecast and other such information.

Apart from information dissemination, we could also provide services such as links to pay for utilities, recharging of metro cards, filing RTIs and more. Given extensive rollout of this in the future, in combination of the extensive mobile penetration in a place like Delhi, this could also be used to conduct public referendums (linked to something like unique identity number (For example VoterID card number) linking them to residency of Delhi. Aadhar has its fair share of problems, so probably not that.).

Most of these are already possible online, but my recommendation would be to have a common startpage or portal come up every time someone connects to the public WiFi, with the portal displaying information and having links to all these services in one central place.

This is not an immediate requirement, but in terms of services, I think what is more essential is access to open data (for non-sensitive categories) in machine readable formats - that is, an API, not PDFs or excel files. Based on this, citizens can make their own apps and services to analyse various metrics of the state's progress. Singapore already does this. To start, we can release real-time or near real-time open data in machine readable format for various pollution measurement parameters like information about the haze in air, ground water levels, etc.

We can also look for Estonia (a country which has embraced the Internet fully) for inspiration on what kind of services they provide online.

5. What is the capex and opex required to create and operate public wifi?

At this point, it is too early to tell. Depending on the kind of implementation which finally happens, this would change. WiFi routers are not very expensive, especially in a project like this where it will be purchased in large numbers (thereby opening ourselves to a heavy discount). It's too early to call, IMO.

Right now, the first thing would be to decide on the right implementation and rollout plan, the budget estimates for each could be discussed at that stage once the proposals become more concrete.

6. What could be a model that makes public wifi viable, sustainable and scalable in terms of usage, technology and financials?

The only way I see it succeeding in the long-term is if it cost-neutral in the long-term. This means this would need to generate a certain amount of revenue each year (if not profit). Advertising and monetisation will be crucial here.

There are various avenues to do that, which can be discussed further. I have certain ideas based on my experience with the browser industry (browsers are also free products, but make money), but they'll need to discussed in length at a later stage.

7. Can public wifi network be made financially self-sustainable if viewed from the welfare service perspective alone?

Not in my opinion. Especially since this government has said that WiFi would be free. Making it paid would also marginalise poorer sections of society, so the government can not charge for it without rolling back on that promise.

In this case, we would have to look for additional monetisation strategies and make it an effective platform for advertising for private players, from which we can generate revenue.

8. Should there be multiple operators or an exclusive operator in a well-defined zone of the city?

Depending things like who owns the cell towers, and other factors, multiple operators can be used. I don't think any promises should be made to any private player offering exclusivity in an area. We've seen time and again that whenever ISP players get any form of exclusivity in an area, quality of service suffers.

I'm sure during negotiations, private players will ask for exclusivity, and we must stand firm on our position on this.

9. What could be the operating architecture that brings all the stake holders together?

A separate organisation would need to be created with proper powers and independent way of working, just like we have with the Delhi Metro Rail Corporation (I think some others have also suggested this, and I fully agree). The task of managing the rollout, maintenance, customer grievance redressal, etc needs to be managed that organisation.

10. What are the advantages and disadvantages of a private enterprise model vis a vis a PSU model?

While a private enterprise model will offer faster decision making and rollout, it may be at the cost of the end user, as the private enterprise's main priority is making money. The claims that private players would provide more innovation and better customer support is suspect given the dismal state of private electricity provider players in Delhi, as well as the general satisfaction with the DMRC.

If a DMRC like organisation is created with it being given a good amount of autonomy and minimum political interference, then that would be best way to approach it, IMO.

Other thoughts

One concern to point would be damage and theft of public infrastructure. Steps would needed to be taken to ensure the routers and repeaters and all associated items are not just maintained and working, but also that its not damaged (either by people or by weather) or stolen. One parameter in judging rollout proposals would need to be this.

If the rollout becomes a success, then it would not just be great for the consumer in just this regard, but will also force network operators and ISPs to up their game, and offer better and faster private services to compete with public WiFi.

Given this, the government should also focus on providing the telecom and ISP industry the means to do so, and offer support and a more conducive atmosphere in terms things like LTE and fibre optic rollout. In the west, the introduction of Google Fibre caused incumbent ISPs in those areas to also provide better service and faster plans - we should anticipate this and the government should do as much as they can do enable that as well, in addition to the public WiFi rollout.