Modern cities are blanketed with Wi-Fi Hotspots. Such network access points can be utilised by Internet of Thing devices to enable wireless sensor network data mining deployments to report to the cloud. Vehicles are such devices that can provide valuable data sensing platforms for both internal operational properties such as speed, fuel and engine heat as well as external environmental features including but not limited to traffic, weather, road conditions and even possible road events such as accidents and approaching emergency services. This paper suggests a network routing resolving method to enable large moving groups of vehicles with embedded devices to route data to fix infrastructural access points. The core of this routing resolving and network healing concept is based on establishing network connections locally without the participation of a controlling arbitrator or centralised router. Further properties put forward focus on objective function for router selection, using properties specific to wireless sensor networks. The proposed routing method LED-WSN (Light weight Edge computed Dynamic Wireless Sensor Network) routing protocol is documented and tested in a simulation environment and on development kit hardware. Assumptions made are listed and possible functionality and drawbacks details. Early concept testing shows encouraging results towards dynamic clustering of vehicular networks and demonstrates lessons learnt in protocol design.