The Indonesia Infrastructure Initiative (Indii) and Pulse Lab Jakarta have just teamed up with Premise to investigate whether we can reliably verify the completion of transport infrastructure projects across Indonesia through crowdsourcing. Below we share our hopes and plans for the experiment.
What links food and roads?
When Pulse Lab Jakarta pitched some of its previous projects to Indii, one stood out: a collaboration with Premise, the World Food Programme and the Food and Agriculture Organisation to crowdsource data on food prices.
Indii liked the look of the model: time stamped and geo-tagged photographs accompanied the price reports shared by users, who, in return, received mobile phone credit or a weekly transfer of funds for their efforts. The right incentives seemed to be in place.
Existing project verification approaches by infrastructure programmes use a random sampling method to deploy engineers. Inevitably a few poor quality roads and irresponsible contractors slip through the net.
Can Indii use a similar crowdsourcing method to the food prices project to verify the completion of its road infrastructure projects? In addition, could this reduce the costs of project verification for a major infrastructure programme?
This is what we hope to understand, eventually, but first we need to establish whether communities are able to generate sufficiently reliable data that engineers based in Jakarta can verify road projects as complete and deem the contractual obligations as satisfied.
Back to Lombok
The current phase of Indii’s road infrastructure work focuses on the beautiful island of Lombok in Nusa Tenggara Barat Province. Thus, the model is being tested on three road infrastructure projects in this location.
The project team is currently deploying the survey via the Premise app to a network of reporters, some of whom are already engaged thanks to the previous project tracking food prices, but others are being engaged through Facebook and other social networks.
Indii produced shapefiles of the road infrastructure projects so that reporters can be directed to the right location for data collection.
The network of reporters are being asked to share information on the condition of nine different infrastructure works, ranging from road surface works to bridge repairs, as well as on drain maintenance works and vegetation handling. The Indii team picked road infrastructure works that non-specialist reporters should be able to identify.
We are learning how best to pose the questions in the survey and what additional information reporters need to generate reliable data, as well as how much we need to pay reporters to incentivise data collection (recognising that these tasks require more effort than reporting on food prices during a visit to a market).
The approach clearly has potential to scale across Indonesia and to significantly reduce the costs of road project verification. We already know that the roads in this initial development and test cycle have been completed to a high standard. But what will engineers unfamiliar with the projects make of the information shared by communities? And how will reporters react to poorer quality road infrastructure works than the current batch on which they are reporting?
We have all this and more to discover during the coming weeks, so watch this space for updates and get in touch if you would like to be involved in the experiment as it develops.