Context and motivation: Software requirements are commonly expressed in natural-language, which must be formalised if they are to be used for formal verification. One application of such formalised requirements is in Runtime Verification (RV), where they can be used to verify that an implementation obeys its requirements at runtime. Question/problem: This paper reports on our experience of using the Formal Requirements Elicitation Tool (FRET) to formalise the requirements for an autonomous tilt-rotor drone in the ProVANT Emergentia research project. Structured, formalised requirements can help to refine the meaning of, and discover ambiguities in, the requirements set, which is beneficial for safety-critical systems such as this. FRET generates a temporal logic semantics for each requirement, providing formulas that can be used by RV tools. Principal ideas/results: We describe the process of formalising the natural-language requirements into fretish, FRET’s requirement language. We present the progress made in each of the four versions of the requirements set as new information was elicited and incorporated, and explain why these design decisions were made. Our two concrete outputs are the fretish requirement set, which can be used in future work for RV; and metrics about the requirements, which help us understand for which RV framework our requirements are suitable. Contribution: From our experience, we present guidance for the FRETish requirements elicitation and formalisation process, which is useful for structured natural-language requirements in general. We highlight situations where it was difficult to express these requirements in FRETish and describe possible improvements to FRET to address these difficulties.