System engineering and science projects: lessons from MeerKAT

Abstract

The Square Kilometre Array (SKA) is a large science project planning to commence construction of the world's largest Radio Telescope after 2018. MeerKAT is one of the precursor projects to the SKA, based on the same site that will host the SKA Mid array in the central Karoo area of South Africa. From the perspective of signal processing hardware development, we analyse the challenges that MeerKAT encountered and extrapolate them to SKA in order to prepare the System Engineering and Project Management methods that could contribute to a successful completion of SKA. Using the MeerKAT Digitiser, Correlator/Beamformer and Time and Frequency Reference Systems as an example, we will trace the risk profile and subtle differences in engineering approaches of these systems over time and show the effects of varying levels of System Engineering rigour on the evolution of their risk profiles. It will be shown that the most rigorous application of System Engineering discipline resulted in the most substantial reduction in risk over time. Since the challenges faced by SKA are not limited to that of MeerKAT, we also look into how that translates to a system development where there is substantial complexity in both the created system as well as the creating system. Since the SKA will be designed and constructed by consortia made up from the ten member countries, there are many additional complexities to the organisation creating the system - a challenge the MeerKAT project did not encounter. Factors outside of engineering, for instance procurement models and political interests, also play a more significant role, and add to the project risks of SKA when compared to MeerKAT.