I got the pleasure to represent Upande at the Mobile Web East Africa, 2010 Seminar; held on the 3rd and 4th of February.

Though meant to be an East African affair, I couldn’t help but notice the absence of our neighbours’. I.e. Rwanda, a fast growing ICT country, Uganda, Burundi and Tanzania and the heavy presence of South African presentations (probably due to MWA09).

Meeting started off with a presentation by Paul Kukubo CEO Kenya ICT Board who were the co sponsors of the event. Paul focused on how their approach is to influence policy formulation, intellectual property, data protection, linkages to venture capital and basically catalyzing growth in the ICT sector between government, the public and business.

Suffice to say the first day was dominated by ‘the big boys’ presentations; namely, The ICT Board, Safaricom, Vodacom SA, Grameen Uganda, AdMob, recently acquired by Google and Rick Jouberts Yonder Media (one of my favorite presentation of the day).

The sessions were quite an eye opener. Laden with statistics, facts and figures, Mobile advertising, Partnerships with Service Providers, Government input, growth and the potentials of Mobile web in Africa. Here are some of them:- 

Africa’s mobile facts

• Africa represents 15% of the world population, but only 3.9″% of the world’s PC internet usage
• Africa’s PC Internet users have increased by 1359% from 2000 to 2009.
• 95% of the African population does not own a Smartphone.
• Nokia 3110c is the most pervasive phone in the market (3.8%), Samsung E250 is at 3.7% penetration.
• Top Smartphones are the Nokia N70 at 10.8% and then Nokia 6300 at 10.3% and then the iPhone at 8.2%.
• African Mobile Web currently has South Africa, Nigeria, Libya, Egypt and Kenya as the top 5.

Kenya's mobile facts

• There were 3.5 million internet users in Kenya last year.
• There are 500 laptops sold per day in Kenya a rise from 20, 000 sales per year.
• Cheapest data enabled devices are about Ksh 2000 in Kenya.
• 70% of mobile data users spend less than 20ksh on a monthly basis.
• Mobile users are estimated at 19.05m subscribers in Kenya with 22m of the 40m Kenyans being the addressable market (15 yrs or older).
• Industry data enabled handsets estimated at 5m or 26% of G.S.M users in Kenya.
• Cost of a 3G handset is 3 times higher than that of a non-data enabled handset.
• Mobile data users estimated to reach 10m in the next five years subject to availability of affordable devices, increase awareness, local content development and drop in data prices.
• If you partner with Safaricom, they’ll take about 60% of revenues meanwhile elsewhere in the  world, like Japan, they give 70% to the developer.

Global Facts

• The global service revenue generated from subscriptions to mobile internet access is forecasted by Informal Telecoms & Media to rise from $57 billion in 2008 to $120 billion in 2013.
• Mobile ad revenue is estimated to be at $2 billion by 2014 and total the value of marketing spent on mobile at around $6 billion.
• Mobile subscription rose from 54 million to almost 350 million between 2003 and 2008.
• On average there is 60% mobile penetration in the world. In developing countries the figure stands at 48%, which is 8x bigger than in 2000.
• Lack of fixed-line access will drive huge mobile internet usage and revenues.
• Vodacom generates 49 million ad impressions per month in South Africa.
• $540 million is the money being made in South Africa in mobile content and the question then becomes whether people can find value in local content.

MWEA10 Day 2

2nd day presentations were more focused on the developers of mobile web applications. 

Started off by a presentation from Mbugua Njihia CEO symbiotic media consortium who cased a couple of their mobile based solutions. 

Eric Hersman ‘Whiteafrican’ followed up next with a beautiful presentation on the Ushahidi. I say beautiful because it has 'location' as a major theme on its Platform. Important to note, Ushahidi are working on a Joomla platform. This would make it easier to use their platform for Websites.

Liko Agosta , founder PesaPal “Kenya’s answer to PayPal” , gave a good account of the application. Tracking back to its inception, from a simple script done from a ‘text editor’ to its complex app it is today. The biggest challenge faced by PesaPal however is Trust! Kenyans being abit weary of using the service. Hopefully Safaricoms M-pesa may have broken that Jinx.

Next presentation was on mobile application and health care in Kenya, from Dr. Kahindi Shedrack.

Steve Vosloo - Shuttleworth foundation gave a good account of E- learning using Mobile phones.

The Other notable presentation was by Jacob Kittinya of Petrol Kenya whose site www.petrol.co.ke is a location based service that lets users know of the gas stations, pump prices and their location in Nairobi. Good use of Google Maps is noted here. Application however is not yet mobile based nor interactive.

Upande is currently looking for an experienced web developer. Someone to help us build high quality web applications. We use the latest technology to create applications of outstanding quality. The ideal candidate must be able to work independently and translate the customer’s wishes into a web application.

It is important that you feel comfortable in a small company with an informal atmosphere and friendly colleagues and that you enjoy keeping tabs on the latest technological developments in the internet field.

Qualifications and experience:

• A higher professional or university education
• At least 3 years of professional experience in the development of web applications using PHP, JavaScript (Ajax), MySQL, XHTML/CSS, Python, Java
• Experience with CMS systems including Joomla
• Exposure to the geo field would be of added value
• Experience with Linux preferred
• Good command of the English language
• Good communication skills
• Flexible work attitude, passionate and committed

What does Upande offer?

Challenging short term initial contract work with the potential for longer term attachment in a small, highly innovative company. Initially internal work with more outward facing roles in the future.
A competitive salary, depending on your qualifications and experience.

Interested?

If you’re interested in working with Upande, please send us your CV including sample work and references by e-mail (info[at]upande.com).

In a previous post on Upande.com, OSM technology used in MapKibera.org was described. One may wonder how the OSM model compares to Google MapMaker which is Google's 'editable Google Maps', and of which all vectors can also be downloaded for free for non commercial use.

OSM differs in the fact that it is mainly built up with GPS on the ground by local experts. This is more accurate though also more time consuming than tracing on top of satellite imagery, which is characteristic of Google's approach. As Google can easily provide high resolution satelite imagery and can benefit from any contributors they are able to mobilize all working in the 'cloud', their approach is more scalable though potentially also less accurate from a thematic perspective, depending on the community they are able to engage.

There are quite a few comparing studies online including Mikel's analysis which analyses coverage for Nairobi objectively but is clearly written from an OSM perspective in terms of licensing. For individuals or organisations wanting themselves to be found through any of the Google platforms within minutes, MapMaker is the way to go. And with Google's dominance in search in general, this is a wise thing to do for most people.

Downloading of MapMaker GIS data is very interesting for the non commercial sectors due to Google's impressive satellite imagery acquisition channels and their unrestricted data licensing model for non commercial use. For full commercial flexibility either in online or offline maps or based on the raw GIS data, OSM has a lot more to offer.

So both approaches have their advantages and disadvantages and have proven that they have made huge impacts. MapKibera definitely fills a hole in any map, including in Google's map. And the level of detail captured in a densely populated area such as Kibera can definitely not be captured on satellite imagery.

MapKibera.org is an initiative by OpenStreetMap's Mikel Maron and Jubal Harpster to map East Africa's biggest informal Settlement in detail. One of the partners is JumpStart who supports social development through strengthening infrastructure and institutions critical to creating employment opportunities and encouraging the development of private enterprise. Their goal is to work with local communities to implement employment, training, and programs that provide working opportunities and promote economic self-sufficiency. They are linked to projects in Gaza, Jordan, Iraq, Jordan, Kenya and the West Bank.

Other partners include SODNET, Carolina for Kibera and Upande.

MapKibera.. Why kibera and why map it...

Though well known, researched and serviced, it remains literally blank on the map.

“Its patterns of traffic, scarce water resources, limited medial facilities, etc. remain invisible to the outside world, and residents themselves. Without basic knowledge of the geography of Kibera it is impossible to have an informed discussion on how to improve the lives of residents of Kibera and other informal settlements as well.”

Due to the level of detail, the maps go beyond just commercial purposes, but are of help to other sectors like Social services, e.g health, local government etc. With more specific data collection, the MapKibera model, as applied world over by Jumpstreet, can be adopted as a basis by the government and NGO’s to frequently gather information.

The longterm plan is to extend projects to other informal settlements and also into formal urban areas and at least cater for the ‘voluntary ‘ part.

Use of the maps...

“Aside from the project’s obvious usefulness for foreign residents or tourists. The maps’ potential for improving local projects, is unlimited: improving opportunities for service delivery, being used for urban planning and logistical purposes and any other infrastructure schemes that arise in the future.”

Methodology...

Kibera has been split into villages (administrative wards) and the mappers are assigned into the wards on the basis of size and familiarity. This approach assures a simple but thorough collection of data. The mappers have basic knowledge of computers and are taken through some training on the use of GPS receivers and the web mapping tool.

Tools in use...

Garmin handheld GPS receivers and OpenStreetMap tools. Data is licensed through Creative Commons Share Alike, meaning it is availalble for commercial use also provided that edits are shared again. Due to the funding data is readily available to general public. The data which is either in OSM or GPX formats is downloadable in other common formats i.e Shapefile and KML. MapKibera aims to produce the first complete free and open map of Kibera, Kibera being perhaps the first in a series.

Technical support and Local point of contact:

“Individuals from the blossoming Nairobi tech scene will help train and make connections with the larger community, and create a sustainable group of map maintainers beyond the initial three week November effort.” SODNET has kindly hosted many of the MapKibera meetings, Upande has provided some technical support on the GIS front.

Online databases such as MySQL and PostgreSQL are very popular for storing large amounts of data online, including geometry fields for storing points, lines and areas. Google has recently launched an online database like product called Fusion Tables which allows one to store large amounts of data in the cloud, analyse the data, link it to other tables, etc. Some really powerful features include viewing the data on a map, charting, views, and export to KML for integration with map products. Until recently one was only able to display point based data using a latitude and longitude coordinate pair, or by geocoding placenames on the fly.

As of this week one is also able to import KML data which represents the different vector formats, points, lines, polygons. This means that Fusion Table's storage can be used for storing large GIS datasets, analyse them (based on attributes) and output them to different map products. See a quick example below of some admin boundaries for Kenya which I imported. 

I will explore these new exciting features more in the (holidays) coming days and show more of what is possible.