Remote handling of breeding blankets poses an unprecedented challenge in future tokamaks like EU-DEMO, where individual blanket segments can weigh upwards of 80 tonnes and extend beyond 10 meters in length. The unparalleled scale of these components, coupled with extremely tight positional tolerances, demands careful consideration of structural flexibility during manoeuvres. This underscores the need for model-based control systems capable of mitigating oscillations and static deflections induced by gravity and disturbances.

In this paper, we present recent experiments towards this objective, focussed on mitigating the effects of gravity. We applied model-based, optimal feedforward control to the motion of a planar slender payload, chosen to represent the flexible part of a candidate breeding blanket segment.  We combined a predictive path planning algorithm with input shaping techniques to manoeuvre the payload effectively accounting for the influence of gravity, while also reducing post-manoeuvre oscillations.  Our results highlight the advantages of model-based control and the limitations associated with the use of feedforward control.