By Mark Ward
Technology Correspondent, BBC News website
Data about people in shanty towns is often old and inaccurate
Handheld computers could help give a voice to the huge numbers of people that do not officially exist.
The gadgets are being used to gather data about the estimated one billion people who live in shanty towns.
The Mobile Metrix project aims to determine how big these communities are and discover what their lives are like.
The up-to-date data will be given to governments and aid workers to help fine tune projects trying to help these communities.
"We count the uncounted," said Melanie Edwards, head of the Mobile Metrix project at Stanford University.
Ms Edwards said many of the people who lived in the world's shanty towns did not feature on government records, and because of that estimates about the size of these communities and the quality of their lives was often old or wildly inaccurate.
One estimate of the people living in one shanty town, or favela, in Brazil put the population at somewhere between 5,000 and 60,000 people. Brazil is thought to have up to 700 favelas.
"If they are off by 55,000 in one community, as you take that up to national or international scale you realise how far we could be off," she said.
"We really do not have a clue yet billions [of dollars] are being thrown in that direction," said Ms Edwards. "These numbers are the ones that a lot of organisations are basing funding decisions upon."
"It's critical because how do you serve that population if you do not know who they are?" she asked.
The data could also help governments commit appropriate resources in the event of an outbreak of a disease such as bird flu.
Aid can be hard to target with no idea of the scale of a problem
The Mobile Metrix project aims to get accurate data about people in shanty towns by employing and training teenagers who live in the communities to go door-to-door and quiz people about their lifestyles.
Using a 100 question survey loaded onto handheld computers the uniformed workers employed by the project find out as much as possible about life living in these communities.
The numbers of teenagers employed in a particular community could vary, said Ms Edwards, depending on whether a sample or blanket coverage was needed.
In one trial project 12 agents using Palm handhelds surveyed 700 families to get a representative sample of the 28,000 people that lived in the area.
In a bid to tempt teenagers to sign up wages for project workers was set to be a little higher than the youngsters could earn from working as a courier for drug dealers.
Data gathering projects were always set up in partnership with leaders in communities and the aim was, said Ms Edwards, to create a training organisation and distribution system that can persist after the data gathering was done.
"Technology is one of the few things that can accelerate things more than most in terms of narrowing the gap," said Ms Edwards.
Mobile Metrix pilot projects have been run in Brazil but others will soon kick off in India and Kenya.