To develop a fully autonomous UAV it is important that it can detect the environmental conditions, so that autonomous Navigation and Guidance can be achieved. Detection of environment can be done using various techniques and broadly can be categorized in two divisions, one is in Radio Frequency environment and another is in Radio Frequency denied environment. In RF environments detection can be done using radio waves based tools such as GPS and Radiolink between the platform and the UAV, however, in RF denied environments, such techniques cannot be used and this work presents a survey of methods using in image and signal processing, carrying out an intelligent approach based on Artificial intelligence and machine learning. Such techniques would be very robust and can work in any kind of environmental and geographical situations. There are many application fields of UAV where radio waves are not available such as in jammed environment, uncovered areas, landing applications in marine environments, surveillance in remote locations etc.

Reinforcement learning based control algorithms require large amount of training. During training the agent explore many possible safe and unsafe states. In the simulation environment exploration of unsafe states can be affordable. But many times, training in simulation world is not a feasible option because of issues of modeling of real-world scenarios. In such scenarios direct real-world training seems to be feasible solution. In such real-world training exploration of unsafe state is very dangerous for the agent as well as for the environment. In this work, safe learning is carried out for multi agent deep deterministic policy gradient (MADDPG) algorithm with the help of control barrier functions (CBF). This MADDPG augmented with CBF is applied for an application of long duration autonomy. Battery driven multiple ground robots are deployed for surveillance of a given region with fix charging stations. The robots are trained for surveillance considering safe learning i.e. robots will not drain out of battery during training and testing for surveillance.

The potential field algorithm which was introduced by Khatib in the 1980s is well-known in path planning for robots. The algorithm is very simple yet provides real-time path planning and is effective to avoid robot’s collision with obstacles and hence it is a reactive path planning algorithm. It can provide the path in a changing environment. The purpose of the paper is to implement an artificial potential path planner in V-REP.VREP is a 3D robot simulation software with an integrated development environment that allows us to model, edit, program, and simulate any robot or robotic system. We have used MATLAB to do the computational aspects for path planning and VREP is used as a visualization platform. In addition to this, we have applied this algorithm to a six DOF drone whose modeling is done in MATLAB. The simulation result we got indicates the fact that the artificial potential field algorithm is a good choice if we used it as a local planner than a global planner. In many cases, the robot got trapped in local minima, which prevents the robot from reaching the target position. Hence It can be concluded that the artificial potential field algorithm needs to be modified to pass all of the local minima problems.