Since the industrial revolution, machines have taken over many tasks from humans, increasing labor productivity and prosperity. In the 20th century, the introduction of robots created a second wave of automation, increasing the labor productivity even further. In order to create a third wave of automation, it is necessary to develop a new generation of robots that is able to act in unknown, unstructured environments, such as households, space and factories in which humans and robots collaborate. Two of the main aspects of robots that have to be improved in order to be successful are their energy consumption and their reliability. This thesis is split into two parts. The first part focuses on reducing the energy consumption of robots by using elasticity. The second part focuses on increasing the reliability by using feedforward control. Throughout this thesis, we focus on one type of robot: robotic arms with repetitive tasks, such as pick-and-place tasks.