Unmanned Aerial Vehicles are becoming more and more popular to meet the demands of increased population and agriculture. Drones equipped with appropriate cameras, sensors and integrating modules will help in achieving easy, efficient, precision agriculture. The proposed solutions related to these drones, if integrated with various Machine Learning and Internet of Things concepts, can help in increasing the scope of further improvement. In this paper, the related work in this field has been highlighted along with proposed solutions that can be integrated into the drone using Raspberry Pi 3 B module.

The wide construct of “biotech” or “biotechnology” encompasses a large vary of procedures for modifying living organisms consistent with human functions, going back to domestication of animals, cultivation of the plants, and “improvements” to those through breeding programs that use artificial choice and conjugation. Fashionable usage additionally includes gene-splicing yet as cell and tissue culture technologies. Researchers are looking for bio-technology management for over a century. Luckily, IoT has the solution to help the employment methodology at every stage of the bio-technology management.

Ubiquitous objects are getting “smarter” and more “connected”, every day. With this ever-growing Internet of Things, every object can now be uniquely identified and made to communicate with each other. This approach has been applied to dustbins too, to monitor garbage collection, throwing light on numerous valuable insights. Our project too employs a similar approach, to not only monitor garbage collection but also optimize it, using machine learning. The method of unsupervised learning we utilize is K Means Clustering, widely used in data mining and analytics. Our physical device uses an ultrasonic sensor to be aware of a dustbin's current content level. If the level reaches or exceeds a threshold percentage of the total capacity of the dustbin, it informs our servers, via an online application programming interface (API) developed for this purpose. The API also stores related data - fill time, cleanup time, and location, to name a few. This dynamic dataset generated is analyzed by our algorithm, to determine the times of the day, when a regular cleanup should be performed, such that the dustbins are clean, for the maximum possible portion of the day. The algorithm also shows the locations, where another dustbin should be installed, for further optimization. This is found out by inspecting each cluster individually and scanning out - items which are the furthest away from its closest centroid; and multiple items related to the same dustbin. In either case, a new dustbin installation is advised at such locations. Data henceforth generated revealed that the installation has had a positive effect on the optimization.