Background
This project is unique because it is a continuation of a program that has been operating for the last six years. Throughout that time, Michigan State University has been involved in implementing technology in various schools in the Arusha region of Tanzania. Projects have ranged from installing solar electricity, to setting up computers, and establishing internet capabilities. There is a need for agricultural technology because in this area, which is inhabited primarily by the Maasai people, resources for irrigation and fertilization are scarce. The Maasai people are traditionally nomadic and have only begun maintaining permanent settlements in recent years. Learning how to farm has been a large adjustment for them, but one they desire greatly. The ability to be able to farm more efficiently could result in significant economic benefits. Initially, several solutions were researched and investigated. There are already numerous options on the market for moisture and nitrogen sensors. Many of them, however, are not wireless. Those that are, are not connected as a network of sensors that analyzes collective data to determine if there is a need for further actions. Research was also done to determine the composition of the soil that sensors will be placed in, as well as the weather elements they must endure. Options were looked into as to the appropriate platform to wirelessly connect the sensors, and the interface to communicate the results. Material researched was evaluated based on the primary constraints of cost, power consumption, and size.
Proposed Solution
The system in Figure 1 depicts a diagram of the proposed design solution. It is essentially a system of sensor nodes that are wirelessly connected to a base station. With receivers on each sensor, as well as on the base, a ZigBee wireless mesh will act as the communication platform to relay all data.
Each sensor node will be placed 12-18 inches deep in the soil, and will contain a moisture sensor and a nitrogen sensor. The moisture sensor will utilize two electrode probes that measure resistance, and convert that resistance into an output voltage. The output voltage will be transmitted to the base station where it will be processed and converted to a moisture percentage.
The sensors will be wired to a microcontroller, and the entire node will be powered by lithium-ion batteries. Each sensor node will be enclosed in a waterproof case, and constructed in a modular fashion. This allows, that in the event of one component failing, it can be replaced without having to replace the entire unit. In an effort for the customer to locate each sensor node, a brightly painted, two-foot rod will extend above the ground. On it will be an ID number that will coincide with a number on the interface back at the base station.
The base station will use a Raspberry Pi as a computer. It will be run through a Linux-based web server and contain a data logger, as well as all system settings. The system will be powered by a 120V AC voltage supply.
Figure 1