Computational optimisation of light distribution in rotating vertical farms: a greenhouse case study

Vertical farming in sunlit greenhouses can reduce energy consumption compared to plant factories, but introduces the challenge of heterogeneous light distribution within vertically stacked cultivation systems. To address this, rotating shelves have been proposed. This study aimed to quantify how strategic management of such a rotating vertical system can optimise light capture and uniformity. We developed and validated a 3D computational ray-tracing model to create a digital twin of a rotating vertical farm in a greenhouse. The model was validated against empirical microclimate data (R 2 = 0.88) and deployed to optimise both static (orientation) and dynamic (rotation) parameters. For the installed system, an alternative orientation enhanced total daily light interception by 3.4%. A separate theoretical optimisation for the isolated system under open skies revealed a significantly higher potential gain of 13.7%, quantifying the constraining effect of the greenhouse superstructure. Furthermore, dynamic management was highly effective: increasing the rotational speed by 33% reduced the spatial coefficient of variation in daily light integral from 46.1% (static) to 3.8%, effectively eliminating positional disparities. These results demonstrate that a shift from heuristic to computational design can substantially improve resource use efficiency. They further establish that while dynamic operational strategies can near-perfectly homogenise light, the full photosynthetic potential of vertical farms is ultimately co-limited by their integration into built environments.