Thrust 1 - Agricultural Sensor Systems

Thrust 1: Agricultural Sensor Systems will design and manufacture resilient, networked, intelligent sensor-robotic systems that monitor the state of plant and soil health over extended areas. Thrust 1 will address fundamental scientific questions to uncover how the complex system of abiotic and biotic variables affect crop yield and resilience, and with this knowledge design technologies and systems that will be deployed with the spatial, temporal, and compositional resolution needed to capture the state of the field. 

Goal

To map spatially and temporally the complex plant and environmental variables affecting crop yield and stress resilience by creating low-cost, zero-/near-zero power sensors distributable throughout the field and remotely interrogatable; to advance robotics and on-board sensor suites for direct biotic crop and abiotic environmental observation and indirect examination of distributed sensors.

Complex System of Abiotic/Biotic Variables

The growth and yield of crops are highly-dependent on the physical and chemical properties of the soil and microclimate where plants are grown. These include abiotic stress conditions such as drought and nutrient deficiencies and biotic stresses related to pests and diseases. However, the inception of stress events is usually highly variable both spatially and temporally. For example, critical soil properties like plant available water usually vary substantially by soil depth and texture, within a field, and within a growing season. This leads to highly variable crop responses that are integrated at a field scale, in the form of impacts on yield and quality, once the crop is harvested.

We aim to understand how the complex system of abiotic/biotic variables affect crop yield and resilience, and on what timescales and with what spatial resolution we should deploy technologies to capture these variables.

Sensors

Thrust 1 will develop and deploy sensors to measure and map the spatial and temporal variability of soil, plant, and environmental properties. We will enhance existing technologies, to make them more accurate, rugged, and affordable. We will create new multi-mode sensor technologies (“smart chaff”) that are low-cost, biocompatible, biodegradable. Such sensors can be widely dispersed close to and directly in contact with plants and beneath the soil surface during the initial planting process or later via non-invasive surface applications. The smart chaff will be small in size (< 1 cm3), and yet large in number to intimately interact with soil and crops across large-area fields to sense the measurands of interest and improve sensing capability and resolution. The smart chaff will require zero or near-zero power, passively and coherently reflecting a remote optical or radio frequency (RF) signal to communicate the state-of-the-field when interrogated by aerial and/or ground-based robots and farming vehicles. We will develop ultralow-cost, large-volume manufacturing processes that leverage methods from printed and conventional electronics to realize low-cost <$0.01-1 sensors. Sensor substrates and packages will be nontoxic and biodegradable so that chaff can be seasonally renewed without contaminating the land or crops.  

Robotic Swarms

IoT4Ag will deploy an autonomous fleet of networked aerial and ground robotic swarms to extract actionable information about row and tree crops in a farm or orchard. Our swarms will enable precise and accurate metric maps that include measurements of every crop (e.g., canopy volume, height, and trunk diameter) at a scale and with an efficiency that has never been accomplished. We will pursue semantic mapping, which includes detection, identification, and classification of relevant objects such as fruit, nuts, trees, pests, and row crops, using deep learning techniques for detection and classification, visual-inertial odometry for localization, and the identified objects for mapping. The swarm will enable us to interrogate low-power, smart chaff sensors by flying low to or driving on the ground using algorithms that will locally maximize information gathering while globally achieving coverage.

Thrust 1 Research