In addition to localization and mapping, current challenges faced by driverless (autonomous) car parks encompass computational complexity, resulting in elevated CPU utilisation and substantial memory consumption. This paper presents a memory-efficient Spike-Time-Dependent Plasticity (STDP) approach for future driverless car park infrastructure (DCPI). The proposed method utilizes self-organized mapping (SOM) classification and memory optimization techniques to analyze traffic occupancy trends. Results from geospatial data mining show optimised memory usage for vacant slots and trajectory mining clusters. Additionally, the impact of lateral suppression on STDP is demonstrated, highlighting the usefulness of STDP in managing complex memory scenarios in DCPI. Results show that STDP provides a robust neuromorphic scheme for lightweight memory utilisation particularly for tasks requiring precise timing and coordination.