604 research outputs found
Long-range collision avoidance for shared space simulation based on social forces
No full textShared space is an innovative approach to improve environments where both pedestrians and vehicles are present, with integrated layouts to balance priority. The Social Force Model (SFM) was used to visualise pedestrian and car trajectories so that peaks of density and pressure at critical locations are avoided. This paper extends the SFM to consider a long-range collision detection and collision resolution strategy. The determination of potential conflicts is enhanced using principle component analysis for a set of agent's prior speeds and directions. This long-range collision avoidance strategy results in more realistic SFM-based trajectories for pedestrians and cars in shared spaces
Environmental ethics: values in and duties to the natural world (summarized with commentary by Panagiotis Perros)
No full textSummarized with commentary in Greek by Panagiotis Perros.Environmental ethics stands on a frontier, as radically theoretical as it is applied. Alone, it asks whether there can be nonhuman objects of duty. Animals, plants, endangered species, ecosystems, and even Earth are progressively unfamiliar as objects of duty, and puzzles arise both for theory and practice. Answers to such questions are as urgent as any humans face, and intimately related to the four principal issues on the world agenda: peace, population, development, and environment
An operational model for liquefied natural gas spot and arbitrage sales
No full textAs more buyers become interested in the spot purchase of liquefied natural gas (LNG), the share of spot trade in LNG business increases. This means that the cash flowing into the upstream of LNG projects is a combination of that generated by deliveries to long-term contract (LTC) customers and uncommitted product and arbitrage spot sales. LTC cash flows are more predictable while uncommitted product and arbitrage cash flows, affected by the dynamics of supply and demand, are more volatile and therefore less predictable. In this research, we formulate an inventory routing problem (IRP) which maximizes the profit of an LNG producer with respect to uncommitted product and arbitrage spot sales, and also LTC deliveries at an operational level. Using the model, the importance of arbitrage, interest rates and compounding frequency in profit maximization, and also the significance of interest rates and fluctuation in spot prices in decision-making for spot sales of uncommitted product are studied. It is proven that writing traditional LTCs with relaxed destination clauses which assist in arbitrage is beneficial to the LNG producer. However, in contrast to what was predicted neither the interest rate nor the compounding frequency has any importance in profit maximization when no change of selling strategy is observed. Apart from these, it is shown that there is a trade-off between the expectation of higher spot prices and the inventory and shipping costs in decision-making for spot sales of uncommitted product in the LNG industry. Finally, it is observed that the interest rate can affect the set of decisions on spot sales of uncommitted product, although the importance of such changes in profit remains to be further explored
Calibration and validation of a shared space model: case study
No full textShared space is an innovative streetscape design that seeks minimum separation between vehicle traffic and pedestrians. Urban design is moving toward space sharing as a means of increasing the community texture of street surroundings. Its unique features aim to balance priorities and allow cars and pedestrians to coexist harmoniously without the need to dictate behavior. There is, however, a need for a simulation tool to model future shared space schemes and to help judge whether they might represent suitable alternatives to traditional street layouts. This paper builds on the authors’ previously published work in which a shared space microscopic mixed traffic model based on the social force model (SFM) was presented, calibrated, and evaluated with data from the shared space link typology of New Road in Brighton, United Kingdom. Here, the goal is to explore the transferability of the authors’ model to a similar shared space typology and investigate the effect of flow and ratio of traffic modes. Data recorded from the shared space scheme of Exhibition Road, London, were collected and analyzed. The flow and speed of cars and segregation between pedestrians and cars are greater on Exhibition Road than on New Road. The rule-based SFM for shared space modeling is calibrated and validated with the real data. On the basis of the results, it can be concluded that shared space schemes are context dependent and that factors such as the infrastructural design of the environment and the flow and speed of pedestrians and vehicles affect the willingness to share space
Capacity Constraints in Public Transportation
No full textThe capacities of public transportation systems are limited in several ways:
Among other limitations, there exist only a finite number of vehicles, space inside
the vehicles is limited, and space inside the stations is limited. In this thesis a transit
assignment model is used, where vehicle capacities are explicitly taken into account
in the strategy choice model. The basic assumption of the model is that passengers
know in advance, which parts of the network will be congested. Passengers take the
possibility of failure to board a vehicle into account before they start their journey.
In the model passengers use strategies instead of routes. A framework for
strategy costs is developed, which is based on random variables. This way it is
possible for the first time to take into account the passenger's averseness to travel
time variability in a public transport assignment model. Furthermore, strategy cost
functions are developed that reflect limited information and bounded rationality of
passengers. Finally, cost functions that reflect the use of portable journey planners
are analyzed.
The assignment model is analyzed in detail on a small bottleneck network. The
results show that the model reacts as expected in all cases. In the model the peak of
passenger arrival times on the origin stop is earlier if there is more demand, which
is a result that is hard to reproduce in models that do not have explicit capacity
constraints. An improved method to model demand is developed. Instead of the
original demand model, which is based on grouping passengers into groups before
the strategy choice is executed, strategy costs are calculated first, and then strategy
choice is executed. As opposed to the original model this method does not suffer from a discretization error and leads to stable results
Performance optimisations for urban last-mile distribution using autonomous vehicle fleets
No full textMany logistics operators are considering deploying autonomous delivery fleets, such as Mobile Parcel Lockers (MPLs) and Autonomous Delivery Robots (ADRs), to address increasing delivery demands and sustainability challenges in cities. These solutions offer economical and efficient responses to urban delivery bottlenecks, such as time spent on parking and manual delivery modes. Regulatory issues, including requirements for using autonomous vehicles on roads and ensuring pedestrian safety with sidewalk ADRs, are major concerns due to varying laws across regions. Additionally, the feasibility, effectiveness, and capacity limits of deploying autonomous delivery modes in traffic-dense cities need validation before large-scale implementation. Considering the envisioned scale of autonomous delivery fleet deployment, conducting feasibility and capacity assessments becomes imperative.
This study developed a macroscopic assessment model incorporating cost, delivery efficiency,
and pollution emissions to predict the performance of a collaborative delivery pattern integrating human-van delivery, MPL delivery, and ADR delivery in different London boroughs. A congestion game model was proposed to explore customer selection behaviour between MPL and ADR services, providing parameters for future deployments. Mixed-integer programming models for MPL and ADR route planning were established, using reinforcement learning methods and graph theory to ensure solution quality and computational efficiency. Geographic datasets from Illinois, US, were used for the ADR route planning study.
Results indicate that MPL and ADR deployment in urban areas reduces operational costs and improves delivery efficiency compared to conventional manual delivery modes. Their lightweight and easy transportability can alleviate urban traffic congestion and parking space occupancy. Rational curb space planning and parking management strategies can facilitate the integration of autonomous delivery vehicles, promoting wider implementation. For logistics operators, differentiated pricing strategies and capacity deployment based on customer preferences, along with vehicle route planning according to customer time windows and activity trajectories, will reduce operational costs and enhance user convenience.Open Acces
Enhancing autonomous vehicle decision-making through scenario-based traffic rule integration
No full textUrban transportation is becoming increasingly complex due to urbanisation, rising traffic density, and growing safety concerns. Autonomous vehicles (AVs) are being developed to mitigate accidents, optimise traffic flow, and improve overall mobility. However, AVs still face challenges navigating complex traffic environments while adhering to region-specific traffic rules. Most existing AV systems lack mechanisms to systematically interpret and enforce traffic laws, particularly in high-risk scenarios.
This thesis presents a scenario-driven framework that integrates traffic rules into AV decision-making by integrating them into reinforcement learning (RL) reward functions. The framework first detects high-risk intersections by analysing aerial imagery, building geometry, and traffic data. It introduces a view percentage metric that links restricted visibility to higher accident rates. The framework builds on DeepLabV3+ or UNet++ and includes additional spatial and visual data to enhance risk prediction.
To enforce compliance, the framework incorporates a Retrieval-Augmented Generation (RAG) system that extracts and structures region-specific traffic regulations for AV training.
To evaluate the effectiveness of this integration, experiments were conducted using a simulation platform under high-risk intersection scenarios. Simulation results demonstrated that AVs trained with traffic rule constraints exhibited substantial safety improvements. Emergency vehicle collisions were eliminated, dropping from an average of 7.36 to 0 per 100 simulation episodes.Open Acces
Investigating the personal exposure to air pollution in London
No full textThe use of fixed air pollution monitors to assess population exposure to pollution may result in significant errors because actual exposure for individuals depends on their activities and distances from emissions source locations. To investigate personal exposure to air pollution, this study: (i) evaluates the accuracy of personal exposure measuring instruments by comparisons to reference instruments; (ii) measures the variation in personal exposure to black carbon (BC) in three different UK cities along different transport modes and routes; (iii) evaluates exposure along a specific walking route in Marylebone, measuring BC and lung deposited surface area (LDSA); and (iv) investigates the spatial distribution of nitrogen dioxide (NO2) measured by diffusion tubes in Marylebone.
BC levels measured by the portable exposure instrument (AE 51) were higher than those measured by the reference instrument by approximately 14% to 16%. BC concentrations recorded on four different transport modes in different cities showed large variation. Walking on quiet streets led to the lowest mean BC exposure when compared with any other route in every city. Children are found to be exposed to higher BC concentrations than adults when walking on the busy streets. On the walking route in Marylebone, BC and LDSA are more highly correlated on busy streets compared to streets with low traffic volumes, which indicate that a greater proportion of the LDSA is BC when closer to transport emissions. Finally, the NO2 measurements indicate that road-side concentrations are higher than urban background values over the entire period.
This thesis indicates that personal exposure cannot be accurately predicted with fixed monitoring stations and that there is significant variability in exposure at different times of the day and along different routes, heavily influenced by transport emissions.Open Acces
Modelling urban street configurations for Autonomous Vehicle flows
No full textDespite the difficulties that have been encountered in the development and deployment of Autonomous Vehicles (AVs) at the current stage, they are still anticipated to become an essential mode of transportation at some point in the coming decades. A growing body of literature has shown that AVs can outperform Human-Driven Vehicles (HDVs) when it comes to fewer driving errors, the ability to predict driving behaviours, and the ability to comply with the increasingly complex traffic regulations in the future. The applications of critical technologies, such as lane-keeping and car-following, could potentially make platoon-based driving viable. That would result in a significant reduction in AV headway, reaction time, and average travel costs, while decreasing the number of road lanes that are required for traffic on the road. Due to the self-cruising and self-parking ability of AVs, novel solutions are being discussed to resolve the paradox of on-street parking problems as a result of the deployment of AV transport systems. A growing body of literature indicates that Shared AV (SAV) mobility will be essential for future Public Transport (PT) services, which are expected to use curbside spaces for parking intensively, frequently visit curbside for Pick-Up and Drop-Off (PUDO) passengers, and for fast-charging operations.
In light of the disruptions brought about by introducing AVs to open roads, more parties from academia, industry, and governments have become increasingly interested in visioning future AV-adaptive streets, which will not only be able to offer AV-based transportation but also encourage the use of various modes of Active Mobility (AM) as well. In order to successfully integrate AV transportation systems into the existing road network, intelligent management over the use of road space - for example, the assignment of road Right-Of-Way (ROWs) and the control of curbside spaces - would be a crucial topic of establishing efficient and AM-aware AV mobility solutions able to interact with the rest of the road traffic.
The present studies contributed to the design of AV-adaptive Complete Streets (CS), however, demonstrating a clear lack of operational and demand-responsive modelling frameworks to ensure that the allocation of road space during the day was coordinated as per the driving, parking, and PUDO demands of AV transport and other forms of transport. To fill these gaps, this thesis proposes three concrete modelling frameworks, aiming to incorporate control over road infrastructure with AV fleet operations.
In concrete terms, the first model designs a Reinforcement Learning (RL)-based model that evolves the configurations of road ROWs and assigns them to respective road users according to pedestrian flow and AV flow in real-time. The objective minimises travel costs and frees up more road space for AM. The second model optimises the layout of curbside parking lanes for SAVs stops in a network by solving a non-linear 0/1 Knapsack Problem (0/1-KP). The third modelling framework trains a RL model to evolve the assignment strategy for the real-time allocation of curbside parking space to diverse types of parking activities, i.e. on-street, PUDO operations and curbside delivery, given a real-world complex network.
To effectively solve these problems, specific metaheuristics and RL algorithms incorporating microscopic traffic simulation have been applied. Various model training strategies were examined, and these models were applied to solve the optima (near optima) under divergent scenarios, to improve the quality of the solution and higher algorithmic performances in respective modelling settings.
According to the results, the proposed method improved the allocation of ROWs in real-time and liberated more room for the use of AM road users with an equivalent to one driving lane width. For the second research problem, the proposed method reduced travel delays by 38.61 s/veh (24.37%) for the SAV fleet compared with respective benchmark conditions. Additionally, the third model balances curbside occupancy with demand-responsive (acceptance rates) to have improved the comprehensive curbside performance by 37.72% from the benchmark scenario. Modelling outcomes also revealed that for multi-agent systems, centralised and distributive training strategies were distinctively beneficial for tackling different problems. Particularly, for the ROWs assignment problem, the distributive strategy was superior, whereas the centralised one achieved better outcomes in the third research problem.
This thesis provides viable Artificial Intelligence (AI)-based tools to plan urban space under considerable uncertainties in the advent of autonomous transportation. The proposed modelling frameworks present novel approaches to the design and management of AV-adaptive streets, which have not been explored in the past. They have the potential to provide feasible and scalable technologies that dynamically assign ROWs to operational AVs and fine-grained curbside parking spaces to accommodate their parking and PUDO demands. Meanwhile, these models allow AVs to safely interact with other road users, especially pedestrians and ensure the efficiency of the transport system under diverse traffic scenarios.Open Acces
- …
