Distributing Industrial Optimization tasks to Rural Worker

Many commentators in academia, government and NGOs have identified the potential of the Internet to distribute work to geographically remote communities. But if such schemes are to make a significant economic contribution to the rural population a long term source of skilled, high value, “distributable” work has to be found. This project is motivated by the possibility that the geometric reasoning tasks associated with many industrial processes and engineering computations have the potential to provide, such a sustained ‘value proposition’ for both buyers and sellers of the work. The belief is based on the observation that humans, regardless of their educational or social background, are adept at manipulating and reasoning about shapes, a task that computers find extraordinarily difficult to do.
The talk will describe on-going work Funded by the ‘Bridging the Urban and Rural Divide’ program the researchers from University of Strathclyde (UoS) in Glasgow, and Indian Institute of Information Technology Allamabad (IIITA) to investigate the thesis that large numbers of industrial optimisation tasks could be outsourced to rural workers to provide a sustainable source of skilled employment.

The goal of the three year program is to deliver a ‘business model’ (supported by the performance and economic evidence from a series of trials) that can deliver long term benefits to the rural population in India and Scotland. The challenges of delivering this objective range from the social and training issues associated with low literacy rates to the establishment of the performance metrics and pricing models required by the industrial customers.

The talk will present some initial results from a series of bench marking trials conducted in Rural Indian Internet Centres and with Scottish Homeworkers to establish how effectively this workforce can carry out tasks such feature recognition, similarity detection and interactive nesting of 2D profiles for sheet metal stamping.