This project aims to develop and validate a novel risk-based tool for planning the reinforcement, replacement, and expansion of distribution grid assets. To this end, we first quantify both known and unknownuncertainties associated with load configuration, including the deployment of electric vehicles, photovoltaics, and heat pumps, as well as the flexibility provided by prosumers. We then propose a multi-stage distributionally robust optimization model that integrates the quantified temporal and spatial uncertainties into a tractable optimization framework to find cost-effective grid asset planning, preferring known risks to unknown ones. The impacts of uncertainties on grid operation are evaluated by accounting for increased voltage levels, overloading of branches, and the aging of transformers. The effectiveness of the proposed method is tested on three real-world case studies, provided by Groupe E, Yverdon Energies, and Services Industriels Lausanne, using appropriate reliability and resilience metrics in plausible scenarios of energy transition.