The Netherlands has positioned itself as a transit nation within the context of military mobility on the European continent. By making its dense and well-developed infrastructure system available, the country facilitates rapid deployment from its seaports to destinations in neigh
...
The Netherlands has positioned itself as a transit nation within the context of military mobility on the European continent. By making its dense and well-developed infrastructure system available, the country facilitates rapid deployment from its seaports to destinations in neighbouring countries. However, using civilian infrastructure for military transits could potentially cause serious disruptions to the travel time of civilians due to capacity limitations. Recognising this tension, the Dutch Ministry of Defence (DMoD) is responsible for balancing the simultaneous use of transportation infrastructure by civil and military vehicles and seeks interventions in the system that can aid them.
The study follows a three-phase methodology:
1. The system analysis examines the transportation system with influences of military mobility by combining insights on the Dutch infrastructure’s characteristics, military movements, and possible intervention strategies for the DMoD.
2. The model development results in a mesoscopic agent-based model (ABM) in NetLogo that simulates simultaneous infrastructure use by civil and military vehicles on synthetic Dutch-like infrastructure networks.
3. The experimentation tests different intervention strategies by simulating their effects on military transit time and travel time loss in the transportation system.
The simulation experiments show that using trains and barge ships is very effective in minimising travel time loss for civilian vehicles, but results in a longer duration of the military transit operation. Combining the road, rail and inland waterway modalities, applying short intervals between vehicles and using long road convoys, are effective in minimising that duration, but cause more travel time loss. The combination of these latter interventions with the avoidance of rush hours and/or daytime balances the trade-off between travel time loss and transit time. Additionally, operating short road convoys should be avoided as they serve neither of the two goals and reserving roads for military vehicles only does not significantly change either of the two goals. This differentiation stresses the importance of setting out the intent of operations - minimising travel time loss or the transit time - before planning them.
The sensitivities and limitations of the research, however, should be kept in mind when interpreting these conclusions. Especially the assumption that the speed flow of highways is capped at 80 km/h on two-lane roads when there are military convoys, strongly affects model outcomes. Other limitations are that the research neglected traffic congestion caused by accidents and road works, and the availability of transport vehicles. Microscopic research on the interaction between civil vehicles and road convoys is suggested to gain insights into these parts of the system.
In addition to partly filling the knowledge gap of simultaneous infrastructure use by military and civilian vehicles, this research also contributes to the scientific fields of managing multi-modal transport and traffic and modelling of transportation systems. Regarding multi-modality, four different management categories were studied, of which three are concluded to be effective: traffic demand management, congestion avoidance, and integrating multiple transport modalities. Prioritisation of roads for military vehicles, however, did not result in significant system-wide benefits in this context.
In the field of transportation modelling, this research confirms the applicability of Agent-Based Modelling (ABM) to this field and introduces a network generation algorithm used before the transportation simulation. The benefits of this algorithm are that it requires less data about the network, which was useful due to the sensitive character of this research topic, as well as a broader applicability of the model, because adjustments in the network generator allow simulation in a different context.