Bottom-up Greenness — An evidence-based greenness scoring platform
The Bottom-up Greenness platform aims to generate real-time activity-weighted greenness scores for any entities in agricultural supply chains backed through data collected at the farm-level. The system consists of three key components: an on-the-ground measurement system at the farm-level, a supervised learning model that can produce predicted greenness scores for a large number of farms, and an aggregation algorithm to generate greenness scores for entities higher up in the supply chain.
The on-the-ground measurement system takes real-time readings of individual production entities such as farmers and micro, small and medium enterprises (MSMEs). Crediting to the collaborative efforts of two FinTech companies – iAPPS and CriAT, a prototype system utilising Internet of Things (IoT) has been developed and is taking real-time readings of more than 200 farmers over 40 defined attributes in Kalimantan, Indonesia.
Real-time readings are then fed into a calibrated model, producing greenness scores for individual farms which are then aggregated to determine greenness scores for entities further up the supply chain. To ensure an arm’s length objectivity and quality of the greenness scores, NGOs will be invited to provide holistic greenness assessments on a sample of farmers. The sample is then used to train a greenness prediction model. Post-calibration, the model can be scaled up to cover a large number of farmers and MSMEs in Kalimantan and elsewhere in Indonesia. To prevent greenwashing and gaming of the IoT system, periodic audits will be conducted on the predictive greenness scores generated by the Model. The audit results will be used in periodic recalibrations of the Model to form a positive feedback loop.