Greenhouse monitoring and harvesting mobile robot with 6DOF manipulator utilizing ROS, inverse kinematics and deep learning models

Abstract

The rapid climate change and scarcity of fertile land has been a global concern recently. To sustain the food supply its high time to think about the modern way of cultivating which is greenhouse. Taking these changes as well as The paradigm shift in people’s occupation, we aim to build a Farming robot with the capability of monitoring and maintaining the soil and the farming environment. In addition this robot will be able to count the amount of vegetables and fruits and harvest them exactly when they are mature for consumption. To move forward with this goal in mind we have added a robotic arm of 6 degrees of freedom and wheel tracks for moving through the mud and soil. With the help of ROS gazebo and A* algorithm rover can make its path through the farm. For picking the vegetables, detecting any diseases on plants we have tried, compared and choose various state of art deep learning models. These models have been merged with object tracking and inverse kinematics algorithms for manipulating the end effector to desired point. Thus, we would have our automated farming robot. The combination of the technologies makes our robot different and effective than other farming robots. As the components used in this robot are easily available and affordable, we hope that this robot would be an active soldier which will sustain our flood supply chain amidst any natural inconvenience.