In this paper we present our works towards creating a natural language platform for an intelligent driving assistant ( IDA) for smart parking in Singapore. In particular, we are focusing on the challenges of designing and implementing reliable spoken dialogue components that enable drivers to communicate hands-free with the system. These components require: spoken language dialogue design, data collection, as well as training of speech recognition (ASR) and natural language understanding (NLU) modules. The main objective of IDA is to help drivers to find suitable parking, online monitor car park availability and redirect drivers when the number of free available spots drops to a critical level. As such, this speech-enabled application contributes to a more sustainable city by decreasing traffic congestion, fuel expenses and time waste for all drivers on the road.