Authors: M. P. Christiansen, P. G. Larsen, R. N. Jørgensen
Multi-disciplinary technologies can be used to exploreand compare design candidates in order to enhance thetime-to-market development for robotic systems. TheCrescendo technology lets software designers and engi-neers collaborate in the development for models con-taining discrete-event (DE) parts of the controller andcontinuous-time (CT) parts for the robot-environment in-teraction. Such models are de?ned as collaborative (co-models) and their joint execution is called a co-simulation.In this paper, we illustrate the development of a robotmink feeding system using the model-based Crescendotechnology. The results of the co-simulations providedan overview of the candidate solutions in the chosen de-sign space entirely in a virtual setting. The candidateoverviews provided valuable input for selecting a candi-date to develop into an actual robot. The selected can-didate solution was subsequently deployed directly on arobot operating system (ROS) based platform and testedon a mink farm.