Degree Type

Creative Component

Semester of Graduation

Summer 2020


Electrical and Computer Engineering

First Major Professor

Manimaran Govindarasu


Master of Science (MS)


Computer Engineering


While there are a lot of precision agriculture data acquisition systems in the market today, not many are equipped with the power of cloud computing. There are a lot of shortcomings of the on-premises solutions, some of which are 1) unavailability of live data globally; 2) unreliable and insecure data management system; 3) lack of scalability due to limited storage availability. Merging cloud computing with the existing Internet of Things(IoT) solutions will not just resolve the above issues but also allow clients to use cloud-hosted applications that will enable them to manage and monitor their system from anywhere in the world.

In this project, we are proposing a cloud infrastructure for data acquisition systems to move the database from on-prem to the cloud and add a plethora of advantages that cloud computing offers. The proposed architecture is based on Message Queue Telemetry Transport (MQTT) protocol and is hosted on Amazon. The key components of this architecture include an MQTT Client, MQTT Broker, MQTT Rules Engine, Cloud-based Database, Server to host client applications, and Client applications (Website and Android). The MQTT Client is the data aggregator sensor node that collects all the information from the field. The MQTT Broker is the bridge between the IoT clients and the cloud and is responsible for routing messages to and from the clients. The MQTT Rules Engine routes these messages to the appropriate destination and takes subsequent actions accordingly. As data is key in any IoT application, all the data is maintained in a centralized Dynamo DB hosted on the cloud. This data is then utilized by the client applications for 1) providing the user with visualization of the information being gathered by the data acquisition system; 2) managing the entities of the system; 3) monitoring the system by setting alerts. For hosting the above applications, we have utilized Amazon’s Elastic Compute Cloud Amazon EC2).

For testing the cloud-infrastructure and the client applications, a virtual Linux machine was set up that emulated the behavior of a CSR-DAQ data aggregator node. This data aggregator node pushed sensor data on the AWS cloud, where it was stored in a DynamoDB table. The consequent test values pushed from the node were reflected in the table in the order they were sent. This confirmed that the cloud infrastructure worked seamlessly and allowed the user to obtain the current status of the soil. The alert mechanism was also validated, and it generated Email and SMS notifications when the soil moisture was not within the user-defined soil moisture limits. Thus, this cloud infrastructure provides a cost-effective solution for data acquisition systems and allows the farmer to access and manage the data from their fields at any point.

Copyright Owner

Mishra, Aditi

File Format


Embargo Period (admin only)