Metaheuristics have a substantial history in fine-tuning machine learning algorithms. They gained tremendous popularity in many application domains. Robotics on the other hand is a wide research discipline that embraces artificial intelligence in a complex individually-thinking robot and distributed robots. Recently, metaheuristics made a significant impact on the application areas of collaborating robotics. This new trend of collaborating robotics, offers the possibility of enhanced task performance, high reliability, low unit complexity and decreased cost over traditional robotic systems. Collaborating robots however are more than just networks of independent agents; they are potentially reconfigurable networks of communicating agents capable of coordinated sensing and interaction with the environment. On the conceptual level, these bots can be empowered by the logics of metaheuristic algorithms which share the same functionalities and capabilities. This paper reviews the recent advances of metaheuristic algorithms on robotics applications. A taxonomy is provided as a reference for robotics designers.