Mapquest, Google Maps or Nokia Navigation have become pretty much defunct standards to travel to any destination. Companies use various technologies to map out specific portions of a community, district, city or country at varying levels of detail to help people navigate.
If you think about it, the existence of a map hints at many achievements for a city, the biggest of which is the fact that there is actually a master plan. The traditional maps of any geographical location are derived from their master plans. Countries in Europe or North America had their master plans available in digital form almost 40 years ago. Their governments plotted every feature on every inch of territory and made the maps available to citizens in paper as well as digital formats. Those governments also made high resolution satellite imagery available to the general public back in the 1970s.
The Navteqs and Teleatlas’s built their sophisticated features on top of these base maps. They saw the void and filled it in as a business. In other words more than 30 years ago, these conglomerates stood on a hill and built their mega structures on top of that hill. Today, a tiny company in Pakistan called Data Solutions, the guys behind naqsha.net, whose only resources are resolve, good intentions, and a bit of know-how, is trying to create that hill to inspire greater, better planned development work.
Nadir Viqar, Team Lead for Software Development at Data Solutions (Private) Limited talks about the value the company brings to Pakistan through its innovation and creation.
What makes these maps different than what Suparco has or what Trakker, Nokia or other businesses that depend on GPS navigation systems?
“The term ‘digital map’ is a generalization used to describe a wide variety of species in the graphics kingdom. A common misconception is that a paper map becomes a digital map if you scan it. Surely, it becomes something digital but not quite a digital map. A scanned map is, in fact, a digital picture. The shapes on it are not objects in themselves but rather representations of the shapes formed by the variation in color, texture, and density. A road in a picture, for example, is simply where the green roughness ends and the grey smoothness begins. But you cannot move the road around, or change its course, width, or length.
If you take it one step further and tell a computer which direction each corner of this picture represents, you will have a ‘geo-coded picture’. Such pictures, like those found on Google Earth, are currently what is being used by companies that depend on GPS navigation. Vehicle tracking can be performed on such pictures since you need to know the location of the vehicle. However, such pictures consume an enormous amount of disk space and can be very slow to load across the internet. Tracking companies, therefore, go a step further by tracing out the shapes in the pictures and storing them as lines and points. This yields a basic digital map which reduces the space and time required to store and load it. Such a map serves the purpose of tracking without going into the finer details.
For example, the roads need not be categorized into large and small, show the direction or speed of travel or even be connected to each other. Nadir explains if you want to be able to enable routing, searching and real-time navigation, a digital map must do the following:
1. Record all lines and points as individual data items in a database
2. Connect all lines that represent roads to accurately reflect the entire road network, including u-turns, roundabouts,ramps, and complex multi-level intersections
3. Assign categories to lines so that a routing program can distinguish between a highway and a street
4. Assign speed limits to all lines so that faster routes can be generated
5. Assign direction to one-way roads so that movement or route of a vehicle can be shown correctly
6. Assign categories to points so that they can be searched or routed to quickly and conveniently
7. Record tens of other attributes for both lines and points to improve the accuracy and quality of information
These 7-points become the cornerstone of our development work, which is what has enabled us to accomplish something great for Pakistan. Each of these features requires meticulous planning and methodical execution to implement. The organizations you mentioned don’t have the need to so meticulous. They could do good business in their areas without incurring costs in terms of time and money.
We’ve been hearing about maps for years, but rarely has a company been able to succeed in the process. What makes your business model different than others?
Our business model and philosophy are inseparable. Data Solutions, the company that owns naqsha.net, was founded in April 2006. The founders have been providing digital mapping solutions to the municipal finance market in the United States since 1997. But the motivation to make maps of Pakistan stems from Naqsha founders’ childhood fascination with maps and travel.
We got into this business because we love making maps, using maps, or just looking at maps, and we wanted to do our bit to pave the way for development for Pakistan. Without maps, development is out of the question. Even something so mundane and grotesque as a housing scheme cannot be planned without first making a map. Yet, knowing this, for 60-odd years our governments have jealously guarded the useless maps they have been making. No wonder we are under-developed and directionless. Without maps, we don’t know where we are going and we will never reach where we want to go.
Quality mapping is Pakistan’s greatest need for planning in the three most crucial areas of development: transportation, health, and education. We ventured into mapping with this intention first. We decided at the outset that the use of the maps would be free for everyone. The business model called for generating quality traffic on naqsha.net so that it should become synonymous with accurate street-level mapping data for Pakistan. We gave ourselves enough funding to build this reputation over a period of two years. We knew that such a reputation would open business doors for us and it did. Since the beginning of this year we have made strategic partnerships that bode well for our future and vindicated the greedless sincerity of our business model.
There is very little data available, at least publicly, for you to build on. How difficult was it to gather the data and track the points you needed to digitize the maps?
In one word, difficult. In two, extremely difficult! When we started planning this project, we didn’t have Google Earth to help us!
Could you describe how you went about making your maps SVG-enabled? We’re looking at vector maps which are scalable and flexible, but also intense in a great deal of detail.
Using SVG (Scalable Vector Graphics) technology was part of an evolutionary process. In the beginning, we struggled with the delicate balance between making the interface look good and having it served quickly. If we improved the graphics, the speed of loading our maps across the internet became too slow. If we sped things up, the graphics started looking jagged. SVG was an excellent solution. However, it provided no data security.
For SVG to work, the map data must be transmitted from the server to the client (your browser). A malicious user could easily capture the data stream and thus “steal” the maps. To counter this, we had to use encryption which slowed us down again even though the performance was within acceptable standards. In the long run, we have decided that it is better not to send the map data to the client. We are therefore working on solutions that will allow use of vector features, such as searching and routing, while serving only images of the map to the client. We make these images scalable by recording them at various zoom levels.
Could you briefly outline the process of how you ensure accuracy of the maps? How are you mapping landmarks and matching them with their GPS coordinates?
Our basic methods and techniques are rather obvious: satellite imagery from a variety of sources, on-location surveys using GPS units of different brands and chip-sets. We are now in the process of acquiring incredibly accurate survey equipment that can pin-point any location within 1 cm of its actual position on the earth’s surface. Going into any further detail would unfortunately require us to reveal some techniques which, for the moment, give us a certain competitive advantage. But suffice to say that we have devised low-cost, high-yield solutions and we’d rather not talk about them for another year or two.
Google Maps has been putting a lot of requests out for online citizens to help tag and map portions of their cities through a GPS tracker. Google, you realize, is a multi-billion dollar company. Don’t you face a lot of skepticism for having achieved what a conglomerate such as Google was unable to really push through?
Naqsha and Google are not attempting the same thing and we’re certainly not competitors. Google is trying to get something done for free. It is a good way to collect an enormous amount of approximate information which would be very useful for many applications. Google has not failed in that attempt and we think it will succeed in the long run. The open source philosophy is one of the greatest contributions of computing to human development and we subscribe to it wholeheartedly.
We hope one day to bring our maps into the public domain and make them freely available to all for input and editing. However, that is not the way to build accurate maps for uses such as real-time navigation. Data integrity is absolutely crucial for such precision applications, and you will tend to compromise on quality when thousands of untrained people are constantly jostling with lines and points.
Once our maps have achieved the level of precision and detail with which we are satisfied, we shall open them up for public input to increase coverage while ensuring that neither our original data is compromised nor new material accepted without checks.
Now that you’ve literally got Pakistan on the map, what kind of community-driven infrastructure do you think would help for your maps to be more interactive and meaningful? (For example: marking locations or consistent objects with RFID tags to ‘coordinate’ a person using your maps through traffic or an optimal route) – have you given any thought to what could happen in the community itself which would make your navigation more effective?
That is obviously every techie’s dream but let’s not get carried away by technology. There are privacy issues here which need to be addressed. Until we become a Republic as proclaimed and the power lies with the people and their rights are inviolable, we cannot go headlong into technologies that will be inevitably monopolized by the enemies of the Republic masquerading as guardians. We don’t dream about RFID tags on people. Our wishes are simpler and humbler.
Having street numbers would be nice. Having postal code boundaries marked on a map would be wonderful! Having house numbers that proceed in some kind of logic that can be reduced to a formula – odd on left, even on right – would be a dream come true. The government can do all that. But will it? If not, can we organize the community to do it on its own? Perhaps, but only after we can organize the community not to pile up hills of rubbish everywhere. Technology can do a lot of things but let’s not engage in ideas that might restrict us more than allow us to grow as people.
What kind of support, if any, have you received from people or companies around you?
For the first two years, none. Most individuals and companies were simply interested in getting their hands on our data for free or stealing it. However, this year we have made some very important strategic partnerships. Our work has been appreciated and we have been given the sort of intellectual property protection space that we absolutely need to continue as a going concern.•
For more details, please visit the company online at www.naqsha.net
