Robocrop is a perfect marriage of IoT and Artificial Intelligence to bring out a solution in the trillion dollar agricultural market. Robocrop enables a plant to move and interact with other IoT devices around it. For Example, If the plants needs water it can move to the location of the tap. It can then turn on the water by directly communicating to the tap therefore realizing pure M2M communication. The soil moisture sensors onboard help the plant to decide when it needs water. Other sensors include luminosity, environment temperature and humidity.
The sentiment analysis model exploits this sensor data to estimate the plant's sentiment and put them out in words (a.k.a slack messages/ text messages). This is done using Microsoft's Text Analytics API which returns the sentiment score of the human stimuli(text message). Microsoft's azure services were used to design the whole system some of which include IoT hub,Stream Analytics, Blob storages, ML studio etc.
As humans we have been negligent of the environment around us. RoboCrop is an attempt to pump love for trees and plants into people. Robocrop makes plants your pets, better, pets you can talk to. Looking at it from a scientific point of view, sentiment analysis provides strong insights on the plant's health. These insights can be used to implement better methods for saving them.
Given the complexity of the entire mechanism and time constraints, deloying a server and loading different modules to interface IoT and A.I. would have been an arduous task. On the other hand, deploying a serverless system was easier and much more intuitive to work with. Why Microsoft though? Power BI marked the pivot point of the decision. While other services provide ML solutions analogous to Azure, A powerful and easy to use UI interface was required to provide real time analytics to users.
RoboCrop can take care of itself. It can calculate amounts of resources needed and when they are needed. This helps in maximizing the yield of the crop.
As a technology, RoboCrop generates 1200 data points per hour which equates to more than 10 million datapoints a year from one plant. The data collected is huge and can be used for further improvements in the trillion dollar agricultural industry. Data is indeed the new currency.
