Nowadays, urban areas are exposed to various challenges such as climate change, social inequalities, and congestion. Shared mobility hubs present the opportunity to reshape our cities and mitigate the previously mentioned challenges by contributing to a more sustainable transport system. These are places where shared cars, mopeds, and e-bikes are offered to improve connectivity in urban areas. In this paper, we investigate the impact of efficiently allocating multimodal shared mobility hubs on modal split, service level, and environmental factors while assuring economic feasibility. Given a limited budget, cities would like to optimize the hubs’ locations to maximize the population's benefits. For that purpose, we introduce a multi-stage design algorithm model that distributes the hubs and allocates fleets of shared cars, mopeds, and e-bikes to maximize travel utility for all the population traveling using traditional and/or shared modes while accounting for multimodal trips. The model is divided into several modules: computational modules that calculate the demand for the hubs; an optimization module to optimize the hubs’ capacities, availability, and relocation of shared vehicles; and finally, a genetic algorithm to find the optimal hub distribution. Our proposed model is one of the first that optimizes the location and capacity of multimodal hubs by considering multimodal trips in a large network. Additionally, it allows to assess mobility, spatial, and environmental impact of shared modes. The model is applied to the case of Amsterdam, the capital of The Netherlands, with around 800,000 inhabitants. After running several scenarios with different budgets allocated to build the hubs, results show that having more hubs with a lower number of shared vehicles is more beneficial than having fewer hubs with higher capacity. That is because the travel time savings increase considerably when investments lead to complete coverage of the area by the hubs network. A modal split of 5% for the shared modes is expected when Amsterdam is covered by 288 hubs. From an environmental point of view, only 32% of the shared trips replace trips previously made by ICE and electric cars, leading to a limited CO2 emissions reduction of 1.27%. Hence, introducing shared modes and mobility hubs without push measures for the use of private cars appears to offer limited benefits to decrease the negative impacts of private car usage.

@en

Freight transporters use software kits to plan the routes for their trucks. Breaks required by the European drivers legislation are nowadays planned as late as possible. However, in some cases it is beneficial to plan these mandatory breaks during waiting time, such as truck driving bans. In this thesis this problem is addressed by computing earliest arrival routes with optimal break planning. The problem is formulated and optimally solved as a variation of Dijkstra's algorithm. The slow Dijkstra computation times of several seconds per route are improved using time-dependent contraction hierachies, which enable a query time of several milliseconds per route while the solution quality remains good. For two days of driving with a night rest in between, 17% of the analysed routes improves with optimal break scheduling, resulting in an average improvement of 5 hours of driving time.

If the taking of breaks is additionally restricted to parking lots, the influence on the arrival time is on average increased with only 3 minutes. However, 5% of the considered routes are not feasible any more due to absence of truck parking lots along the planned route. Another 15% of the routes face large changes in roads that should be taken.

The planned routes are all optimized for having the earliest arrival time. However, the freight company's objective also concerns fuel optimized driving or providing reliable arrival times to the customer. This thesis analyses such preferences and combines them into a route planner that incorporates results of stated and revealed choice experiments. It is shown that generally toll avoidance and congestion avoidance have most influence on the route choice and arrival time. On average differences are small, but for some routes this leads to changes of up to 15% in travel time.

This thesis analyses the problem of route planning from two perspectives: algorithmic and behavioural. Several observations are made: algorithms are not suitable for obtaining the best route instead of the fastest, and the output of behavioural research cannot be used directly in practice to compute preferred routes. Effort could be made in future to integrate both research communities such that algorithms are able to reflect what a freight transporter actually wants.