Intelligent mobility planning for a cost-effective object follower mobile robotic system with obstacle avoidance using robot vision and deep learning

Few industries use manually controlled robots or automobiles to carry material to the desired position, and in some cases, man power are used due to a lack of money. This cannot be used all the time, in all places and all conditions. So, it is very tranquil to have robots which can follow a specific human by following the unique coloured object held by that person. So, we propose a robotic system that uses robot vision and deep learning to get the required linear and angular velocities, which are ν and ω, respectively, making the robot avoid static and dynamic obstacles when following the unique coloured object. We propose a novel LSTM cell called TF-LSTM, which makes the proposed methodology very accurate in tracking the object in 3D space. TF-LSTMs or target follower LSTMs are inspired by the traditional LSTMs and give a meagre error in linear and angular velocity prediction. The PI controller, which was used to control the linear and angular velocities, which in turn controls the position of the robot, gave us impressive results, and this methodology outperforms all other methodologies for precise target tracking in performance comparision. The proposed TF-LSTM gave us an accuracy of 96.1%, average linear jerk of 0.4 m/s 3 , average angular jerk of 30 degrees/s 3 , average clearance of 0.514 m, maintaining no collosions with the obstacles.