Discussion Report 29 Christian Kreutz & Pete Cranston - Using mobile phones for knowledge sharing

From KM4Dev Wiki
Revision as of 19:03, 1 December 2010 by Davide Piga (Talk | contribs)

(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search


Using mobile phones for knowledge sharing

== Convenor == Christian Kreutz & Pete Cranston

== Participants == Christian Kreutz, Pete Cranston, Verele De Vreede, Margarita Salas, Gauri Salokhe, Sabrina Erné, James Marchant, Michael Roberts, Luca Servo, Pete Shelton, Taline Haytayan, Danny Aerts, Damir Simunic, Carl Jackson and Peter J Bury.

== Key Discussion Points ==

Podcast with Pete Cranston on using mobile phones for knowledge sharing: http://www.switchpod.com//users/talinehay/PeteCranstonPodcast.mov

Podcast with Christian Kreutz on mobiles including recent Kenya Crisis http://www.wkg.uk.net/KM4DEV08/Christian_K.mp3 (2MB) - streaming version coming soon

Blog post about the session by Christian Kreutz: http://www.crisscrossed.net/2008/07/20/mobile-phones-for-development-grassroot-innovations/

Blog post with a list of mobile application for development: http://talksharelearn.wordpress.com/2008/09/12/who-is-using-mobile-phones-for-development/

Notes here below are additions from KM4Dev Discussion forum:

Mobile Phones for Social Change in Africa - a talk given by Dr Gary Marsden in March 2008, University of Cape Town

"Mobile interaction design for developing nations" or "People are people, but technology is not technology"

Gary Marsden is an Associate Professor in the Department of Computer Science at the University of Cape Town in South Africa. With a background of academic work in the UK on problems of computer usability and mobile computing, since moving to Africa he has focused on the design of digital technology for developing world contexts, and his current emphasis is on education and healthcare systems based on cellular telephone handset technology.

Not many people in Africa have access to the Internet: only 4.7% of the population, compared to the world average of 20%. Looking more widely across the developing world, we find that 10% of the population in those countries have access to the Internet, compared to about 60% in the developed world.

However, the proportion of people in developing nations who have a cellular handset is much higher: about one third. And Africa is the fastest-growing cellular market in the world, growing by about 40% year on year. What is more, the network coverage is surprisingly extensive, as Gary demonstrated, showing a chart of the 43 nations in Africa documenting the proportion of the population who live in a cellphone-enabled area. For example, Zimbabwe, in dire economic trouble with 26,000% inflation, has 50% of its population within a cellphone access area. (Note: there are more than 150 million cellphone subscribers in sub-Saharan Africa, but a further 350 million have cellphone coverage without having their own handset. - CT)

Why has cellular been so successful? Gary showed the audience a photograph of an item he had bought from a roadside vendor: a colourful bowl woven out of stolen trunk cable, which once had connected a fixed phone line subscriber to the exchange. Cable theft is a costly headache, and the cause of widespread system failure for telecomms operators across Africa...

The Western Cape region of South Africa is the second largest exporter of copper to China, the rade valued at 77 million rands (US $11 million) in 2006, without having a single copper mine. Replacing stolen cables in South Africa costs 500 million rands annually. (See article in Taipei Times: http://www.taipeitimes.com/News/world/archives/2007/06/21/2003366202)

 There are also sociological reasons that favour the cellular
 network.  If you are an itinerant or migrant worker, you can't
 have a fixed line, but you can have a mobile.  You are not
 dependent on electricity either, which has become a big problem
 in South Africa recently.  Cellphones are relatively robust,
 and because they are subsidised by the networks, they are cheap.

 The attachment of Africans to their cellphones is remarkable.
 Gary showed us a photo of an Internet cafe in north west Zambia,
 a remote region bordering DR Congo and Angola.  The Internet
 cafe is collapsing.  Why?  Because there is no cash in the
 village.  The village got cellphone coverage four month ago,
 since when everyone is spending their cash on cellphone airtime.
 The Internet cafe closed for lack of business; the grocery is
 teetering on the brink of survival; even the pub is suffering.
 Gary said that he had met people who would skip meals in order
 to afford airtime.

 In Africa, ICT = cellphones

 When Gary reads articles about ubiquitous computing that
 come from a Western, developed-country perspective, they
 talk about *depth* -- increased density of computers in a
 given area.  He asked his audience to think about *breadth*
 -- spreading computers and computer-like devices throughout
 the world.  Forget the PC; even forget the $100 laptop --
 Gary showed us a photo of the Motorola C113a $20 mobile
 phone commissioned by the GSM Association with the aim
 of reaching "the next billion users", and its follow-up,
 the Motofone F3, with an E-ink electrophoretic "electronic
 paper" screen, giving extremely long battery life on standby.

 For these cellphone owners in developing countries, the
 cellphone is the first and only piece of technology in their
 lives; and they relate to it very differently to how cellphone
 users in the rich countries would.

 Gary owned up to being a white, middle-class, middle-aged
 nerdy engineering guy who loves technology; and he has
 loads of experience designing technology for other white,
 middle-class, nerdy people.  However, the discipline of
 HCI (Human-Computer Interaction) introduces humility;
 and he has learned a lot of humility since moving to
 Africa. HCI methodology directs us to create technology
 based on careful observations of what people want and
 what they are doing; and then evaluating the technology
 in the light of whether it serves those users.

 This means we may have a significant amount of unlearning
 to do.  Consider this: each of us at this conference has
 a cellphone, and it is that person's own cellphone. In
 South Africa, many people use shared cellphones, such as
 arte accessed through a phonecall-centre business like
 the one of which Gary showed us a photo, based in a goods

 The papers given at this Royal Society conference about
 trust and authentication all made the assumption that the
 devices in question belonged to just one person; but that
 would not be true in the townships...

 African cellphone users can be fantastically price-
 sensitive.  While in Zambia making some observations,
 he saw a guy receive a call, look at the number as it
 flashed up on the screen, and immediately switch the
 phone off.  He then extracted the SIM card, inserted
 another, turned it on and made a call...  On asking
 this man for an explanation, Gary learned that of the
 three cellphone service providers in Zambia, it is
 cheap to make local calls on Celtel; MTN has the
 biggest coverage, and CellZ is free to receive calls
 on.  So, you keep your CellZ SIM card in most of
 the time, but you swap to CelTel to make a call to
 a landline.  In the West, we could not be bothered
 for the sake of saving a few pence.  In fact these
 users could do with having phones with at least
 three SIM cards in them.

 In Africa, the cellphone is the computer

 For many people, the cellphone is the only computing
 device they own.  If they need a calculator, they
 will use their cellphone.  You can even see employees
 in the banks doing their calculations on cellphones.

 Camera-phone usage differs, too.  Imagine that the only
 camera you have ever owned is your camera-phone; and
 that you don't have the facilities to transfer photos
 from your phone to a PC, or put them on an Internet
 social networking site, or get them printed out.
 Therefore, for many Africans, their phone is their
 photo album.  They have their wedding photos on
 their phones.  Their phones contain a pictorial
 documentary of their lives.  Getting to know
 someone, they may pull out their phone and say,
 Look, this is my wife; this is my house; these
 are my kids...

 So, one project they implemented at the University
 of Cape Town was a system that makes it easier to
 browse photos on a cellphone.  They got the phone
 to broadcast the image to the screens of other phones
 in the vicinity, to help you show your photographs
 to your friends.  People liked that, because it
 fitted into their culture in a very natural way.

 Most school-children in South Africa use a system
 called "MXit" (see http://www.mxit.com/web/whatis.htm)
 MXit is a basic Internet chat application for the mobile
 phone, and five million people use it; because in South
 Africa, the cost of sending a single character via MXit
 is one ten-thousandth of the cost of sending a single
 character via SMS.  For two rand a day, less than 20p,
 these kids can stay all day on MXit, despite the fact
 that it has a terrible user interface that the likes
 of us wouldn't put up with.

 Many of the schools have banned use of MXit.  But Gary
 and his colleagues discovered that the kids use MXit
 to do their homework collaboratively.  Therefore,
 they added functionality to the MXit system, having
 reverse-engineered the protocol, and added these
 features and functions into some of the chatrooms.
 The kids loved it.  Remember, they have no Internet
 access.  They added an equation-solver, for solving
 quadratic and linear equations, and an interface to

 Harnessing the cellphone for healthcare delivery

 Gary introduced another of the projects his group is
 involved with, a healthcare company called Cell-life
 that was set up to address South Africa's serious HIV
 problem with electronic health technologies.  Cell-life
 (http://www.cell-life.org) uses mobile phone technologies
 in many innovative ways, e.g. to provide a free information
 service via SMS, and to let people book clinic appointments.
 If a person is diagnosed HIV-positive, they are interviewed
 for acceptance into an AntiRetroviral treatment programme
 and assigned a therapeutic counsellor.  These ladies liaise
 with 15-20 patients, collecting essential information on
 the basis of which a doctor will make a prescription and
 a care manager will direct a programme of care.

 Various approaches to information-gathering were tried,
 and met with problems, until someone noticed that all the
 health visitors had their own cellphones.  So Gary's team
 translated all the forms onto the cellphone: forms about
 adherence to treatment, about symptoms, about appointments...
 these generated SMS messages.  Of course the screen is so
 small, and it might have been nicer to deploy the system
 on a handheld computer... but the health visitors were
 comfortable with their mobile phones, and this ensured
 a very enthusiastic acceptance.

 But the University of Cape Town team has had its failures
 too.  Another healthcare-related problem which they tried
 to tackle was in the Eastern Cape, KwaZulu Natal, a very poor
 area, where there was a 100-bed hospital with ten outlying
 clinics that were affiliated to it: and for all of this,
 there was only one doctor.  (That doctor was Cuban: South
 African doctors have been emigrating in droves to work in the
 British National Health Service where the money is better.)

 The team found that nurses would frequently come across
 a condition which they did not recognise, and where the
 opinion of a doctor was required.  For example, suppose
 a man had come in with a purple, swollen leg.  He would
 have to be told to take a taxi over to Mtata, and then
 another taxi to the hospital, all of which he could not
 afford...  So the team thought, why not put in a wireless
 network, and the nurse could take a digital photo of the
 purple leg, and send it to the doctor; and they managed
 to get some aid funding for this, and built the system.
 There was even a nice feature whereby when one of these
 requests for diagnosis arrived on the doctor's PC he
 would be alerted via SMS.

 The nurses had never used a computer before, and it took
 six months to train them.  As good HCI practioners, of
 course they did an evaluation of the system and they
 did find there were some problems with it, so they set
 about making improvements.  And at least one nurse,
 in Gary's words, "freaked out".  "Don't change it!"
 she pleaded.  "But it's very hard to use..."  "No,
 I have spent six months learning how to use this system
 -- do not change it!"  Hmm, they thought, that's not
 what it says in the HCI textbooks...

 Once again, they decided to try porting the application
 onto cellphones -- Windows cellphones using the .NET
 Compact Framework.  And this time, the nurses didn't
 need any training.  The irony was, it was exactly the
 same piece of software, but they found it easier to use
 on the phone.

 The problem was: they never ever used the system.  Why?
 After two months of study, and a lot of interviews,
 they discovered that nurses do not talk to doctors.
 That is not how their social structure operates.
 The technological "fix" which had seemed such a good
 idea was, in that context, useless; and all they had
 succeeded in doing was to upset the nurses.

 Finding & using the "bridge" person

 There is an approach to interface design called User
 Centred Design, in which the user and developer work
 together as co-creators of the technology, right from
 the start.  It sounds like a wonderful idea, but they
 are finding that it doesn't work in Africa.  Because
 that methodology assumes that the user of the technology
 understands it on some level.

 So, the people that Gary and his colleagues often deal
 with know nothing about the technology; on the other hand,
 people like Gary know next to nothing about the culture
 of the community, which really gets shown up when the
 application operates in the context of a social structure.
 The gap is just so great that often we don't even know
 what are the right questions to ask.  Besides which,
 there are just not enough people doing the kind of
 application design work that Gary does.  Consider
 the Cell-life information gathering project: every
 clinic in Africa which had read about that system
 phoned up and wanted a version for their clinic.
 There was just no way they could have done that.

 So now, they have a new way of working, and he suspects
 it is the only way to make computing ubiquitous in the
 developing world.  They have found that within each
 community they can identify a "bridge person", a human
 access point.  Consider, for example, the owner of that
 Internet cafe in north west Zambia, Roy.  He's lived
 in that village all his life, but he got some kind of
 scholarship and received training about IT and the
 Internet.  Now Roy can understand the computing world,
 and the world of the African village.

 So now, Gary's team works with people who can act as a
 "bridge"; they create software with them and for them,
 rather than for the end users, and they make sure that
 the software is sufficiently easy to customise, so that
 the "bridge person", the "human access point", is
 empowered to support the end users.

 So, for example, for a development of the idea of using
 cellphones to capture health-related information, one
 of Gary's research students interviewed 150 doctors,
 got them to sketch out how they would embody the task
 of information capture, and then embodied that in his
 software.  The development system that they used took
 these electronically-sketched ideas for interaction
 screens, and turned them (by clicking on the "Deploy"
 button) into an actual application ready to be burned
 into the SIM cards of the users: it's just a matter of
 giving the application the mobile phone numbers of the
 phones onto which the application should be burned.