Interconnected Pathways to Development

Lead PI: Dr. Vijay Modi

Unit Affiliation: Quadracci Sustainable Engineering Lab

September 2021 - August 2024
Active
Global
Project Type: Research

DESCRIPTION: Literature especially in developing country settings goes back several decades on the attempt to understand the association between energy consumption and socio-economic wellbeing. Understanding this association is crucial to understand the affordability question, the appliance ownership constraint and the question of targeting subsidies. Associations with lower power consuming devices also has implications for electricity planning. In the past the data for such work was limited to a very coarse resolution obtained from surveys at administrative unit level. With the heterogeneous nature of growth (i.e. emerging wealth of a small fraction of households amongst otherwise poorer areas), may now possible to establish. Our recent work suggests that it is possible to utilize utility electricity consumption data of a subset of grid-connected customers to inform the consumption levels (our current work allows us to classify amongst high and low), We intend to supplement our satellite imagery-based estimations with a limited on the ground surveys of household characteristics and appliance ownership to establish these associations.

Support ground data collection methods, sensors and tools:
We intend to vehicle-mount a pair of cameras with GPS capabilities and where needed a lean instrument box (e.g., an accelerometer) that can be battery operated to be able to at a minimum record a time stamp, vehicle location, along with 4K-level stereo imagery on either side of the vehicle to be able to capture visual detail video within a +/- 250 meters buffer of vehicle travel. The primary goal of this effort is to provide a video-stream that can then be processed to create label data under controlled conditions, e.g. when during the year data are collected and with controlled geography for sampling. This dataset will allow us to label farming plots, their irrigation status, and potentially: land-use, building structures/roof materials and other potential indicators of wealth/income, power lines and transport infrastructure. We conceive a hierarchical approach where lower-cost approaches such as labelling of high-resolution imagery and labels from partners (from Atlas-AI effort) will be the preferred first approach which is then supplemented with field-gathered data that might be limited in volume but high in label accuracy and in some cases impossible to gather otherwise.

The key enabler of dry season agriculture and supplementary irrigation is energy. Dry season agriculture can be both a necessity for food security, an improved livelihood measure- potentially adding higher-value horticulture and tree crops to the income stream and enhanced nutrition. The domestic in-country demand and export markets for such crops are also growing with urbanization and economic growth. With the near exclusive focus on household electrification, the emergence of such activities is missed by energy planners. Being able to predict existing energy use granularly is key to enabling service provision, enabling private sector to identify customers for lower operating costs with financing. Projecting the potential future scale is can also enable focus on clusters for agriculture support, transport and logistics planning, and to energy service providers. We will use a combination of satellite imagery, labeling and ground truth to develop analytics applicable to small-holder settings.

Understand the association between energy consumption and socio-economic wellbeing. Understanding this association is crucial to understand the affordability question, the appliance ownership constraint and the question of targeting subsidies. Associations with lower power consuming devices also has implications for electricity planning. In the past the data for such work was limited to a very coarse resolution obtained from surveys at administrative unit level. With the heterogeneous nature of growth (i.e. emerging wealth of a small fraction of households amongst otherwise poorer areas), may now possible to establish. Our recent work (across UMass, CU, RIT) suggests that it is possible to utilize utility electricity consumption data of a subset of grid-connected customers to inform the consumption levels (our current work allows us to classify amongst high and low), We intend to supplement our satellite imagery-based estimations with a limited on the ground surveys of household characteristics and appliance ownership to establish these associations.

SPONSOR:

University of Massachusetts Amherst

ORIGINATING SPONSOR:

Rockefeller Foundation

KEYWORDS

development developing countries

THEMES

Stewardship of the planet Sustainable living