While science fiction often portrays robots with incredible capabilities and versatility, advancements in artificial intelligence, machine learning, and robotics are gradually enabling the development of sophisticated machines that can perform an array of tasks across diverse domains. These innovations have significant potential to revolutionize industries, enhance human capabilities, and shape our future society; however, it is essential to approach these developments with caution and ethical considerations in mind. The caution is on technical feasibility in terms of Verification, Validation, Certification, Acceptance Testing, and Trust. Together with ethical considerations, a design space is generated: the space of all parameter settings for all options a designer has in order to design a robot performing a specific task. In an engineering process, the specific task is formulated as a functional requirement, and the interesting work for a robot designer is to find an optimal operational setup within the design space to implement a robot performing the given task. If the given task is rather vaguely formulated, the robot needs to handle everything that is not explicitly preplanned, in an autonomous manner. Now, if the robot handles situations on its own, how can the designer then be sure that it handles everything correctly? What is the constraint for a robust and efficient autonomy in the design space? Let’s explore together this area in the design space: Where is the robust and efficient autonomy region? How can we measure how far we are away from there in our current design? How can we formulate constraints in our design space such that the autonomy is working with sufficient robustness and efficiency?