Autonomous cooperative transport has been a receptive research hotspot for decades. The main concern is to find the effective coordination among autonomous agents to perform the task in order to achieve overall performance accuracy. In this paper, a multi-robot system performing a payload manipulation task where the participating robots do not have any knowledge regarding other robots in the system or the payload dynamics is considered. A decentralized proportional-derivative control strategy by which the multi-robot system can organize itself and adapt dynamically to changing circumstances is proposed. The decentralized controller is augmented with an orientation control by which the payload in transport attains a desired angular orientation at the target location. An obstacle avoidance strategy using an artificial potential field method is incorporated to avoid any obstacle in its transport path. The transport systemis simulated with uniform and non-uniform distributions of robots around the payload, for transportation as well as obstacle avoidance scenarios.