This paper addresses the challenges of conceptualising, modelling, and planning for the future of complex systems with diverse physical, software, and/or cyber components during feasibility analysis. Specifically, the challenges dealt with included how to identify features; how to model the system, synergies, and operations upfront; and how to predict aspects such as the performance, assembly time, and cost in the fuzzy front-end stage of systems development processes. We present the techniques and scheme we developed and illustrate their applicability by using a practical example of complex medical rehabilitation device. We show that they provide a systematic way of identifying components and features. We discuss how features can be used as the basis for building functional, structural, or full-fledged system model upfront. These models show which functional and non-functional features the system would be comprised of, and can therefore also be used as the basis for investigating flows of information, energy, and/or materials, and for making predictions and planning for the future.Keywords: complex systems; conceptualisation; modelling; prediction.Reference to this paper should be made as follows: Opiyo, E.Z. (2017) 'A feature-based approach to conceptualisation, upfront modelling, and planning for the future of complex systems', Int.