Project

Integrated Cooperative Automated Vehicles (i-CAVE)

Goal: Address current transportation challenges regarding throughput and safety with an integrated approach to automated and cooperative driving. In i-CAVE, a Cooperative Dual Mode Automated Transport (C-DMAT) system is researched and designed, consisting of dual mode vehicles which can be driven automatically and manually to allow maximum flexibility. The program integrates technological road maps for automated and cooperative driving, accelerating the development of novel transportation systems addressing today and future mobility demands.

Access https://i-cave.nl/ for more information.

Updates
0 new
0
Recommendations
0 new
0
Followers
0 new
11
Reads
1 new
129

Project log

Frederik Schulte
added a research item
Carriers can remarkably reduce transportation costs and emissions when they collaborate, for example through a platform. Such gains, however, have only been investigated for relatively small problem instances with low numbers of carriers. We develop auction-based methods for large-scale dynamic collaborative pickup and delivery problems, combining techniques of multi-agent systems and combinatorial auctions. We evaluate our approach in terms of both solution quality and possibilities of strategic behaviour using a real-world data set of over 12,000 orders. Hence, this study is (to the best of our knowledge) the first to assess the benefits of large-scale carrier cooperation and to propose an approach for it. First, we use iterative single-order auctions to investigate possible collaboration gains for increasing numbers of carriers. Our results show that travel costs can be reduced by up to 77% when 1000 carriers collaborate, largely increasing the gains that were previously observed in smaller-scale collaboration. We also ensure that individual rationality is guaranteed in each auction. Next, we compare this approach of multiple local auctions with an established central combinatorial auction mechanism and observe that the proposed approach performs better on large-scale instances. Furthermore, to improve solution quality, we integrate the two approaches by allowing small bundle auctions in the multi-agent system. We analyze the circumstances under which bundling is beneficial in a large-scale decentralized system and demonstrate that travel cost gains of up to 13% can be obtained for 1000 carriers. Finally, we investigate whether the system is vulnerable to cheating: we show that misrepresentation of true values by individual participants sometimes can benefit them at the cost of the collective. Although such strategic behaviour is not straightforward, we also discuss different means to prevent it.
Franz Lampel
added 2 research items
In orthogonal time frequency space (OTFS) modulation , information-carrying symbols reside in the delay-Doppler (DD) domain. By operating in the DD domain, an appealing property for communication arises: time-frequency (TF) disper-sive channels encountered in high mobility environments become time-invariant. The time-invariance of the channel in the DD domain enables efficient equalizers for time-frequency dispersive channels. In this paper, we propose an OTFS system based on the discrete Zak transform. We show that the presented formulation simplifies the derivation and analysis of the input-output relation of TF dispersive channel in the DD domain.
Phase-coded frequency modulated continuous wave (PC-FMCW) radars represent an extension to frequency modulated continuous wave radars. PC-FMCW radars embed a phase code (PC) into the transmitted signal that allows, for example, to suppress interference from other radars or to realize joint communication and radar systems. At the receiver, the PC is removed from the radar signal before the received signal is processed. Therefore a frequency-dependent time shift is applied. This frequency-dependent time shift, however, causes dispersion of the embedded PC. Consequently, a residual error is introduced, and the radar performance is degraded. The dispersion of the PC has been neglected so far in the literature. In this work, we analyze the dispersion of the embedded PC and show that it is independent of the intermediate frequency of the signal. Based on this observation, we propose two novel techniques that allow for the compensation of dispersion. We demonstrate the effectiveness of the proposed techniques using simulations.
Frederik Schulte
added a research item
With the popularization of transportation network companies (TNCs) (e.g., Uber, Lyft) and the rise of autonomous vehicles (AVs), even major car manufacturers are increasingly considering themselves as autonomous mobility-on-demand (AMoD) providers rather than individual vehicle sellers. However, matching the convenience of owning a vehicle requires providing consistent service quality, taking into account individual expectations. Typically, different classes of users have different service quality (SQ) expectations in terms of responsiveness, reliability, and privacy. Nonetheless, AMoD systems presented in the literature do not enable active control of service quality in the short term, especially in light of unusual demand patterns, sometimes allowing extensive delays and user rejections. This study proposes a method to control the daily operations of an AMoD system that uses the SQ expectations of heterogeneous user classes to dynamically distribute service quality among riders. Additionally, we consider an elastic vehicle supply, that is, privately-owned freelance AVs (FAVs) can be hired on short notice to help providers meeting user service-level expectations. We formalize the problem as the dial-a-ride problem with service quality contracts (DARP-SQC) and propose a multi-objective matheuristic to address real-world requests from Manhattan, New York City. Applying the proposed service-level constraints, we improve user satisfaction (in terms of reached service-level expectations) by 53% on average compared to conventional ridesharing systems, even without hiring additional vehicles. By deploying service-quality-oriented on-demand hiring, our hierarchical optimization approach allows providers to adequately cater to each segment of the customer base without necessarily owning large fleets.
Robbin van Hoek
added a research item
Vehicle automation has become an important topic in recent years. It is aimed towards mitigating driver-induced traffic accidents, improving the road capacity of the existing infrastructure as well as reducing fuel consumption. Two major classes of automated vehicles can be distinguished. The first is the class of cooperative vehicles, which use vehicle-to-vehicle (V2V) communication, or vehicle-to-infrastructure (V2I) communication in order to exchange motion data and sometimes meta data such as intention. An example of cooperative vehicles that use inter-vehicle communication is Cooperative Adaptive Cruise Control (CACC), in which vehicles drive at very small inter-vehicle distances and use communicated inputs from preceding vehicles in order to maintain a desired spacing. The spacing policy and controllers of these vehicles aim for string stability, which is the property that disturbances attenuate in upstream direction of the vehicle platoon. This is pivotal in preventing traffic jams and thus for increasing road capacity. However, due to the pre-defined behaviour of vehicle following, the vehicle is only capable of driving in a limited number of scenarios. The second class concerns autonomous vehicles. This type of vehicle uses on-board sensors such as radar, LIDAR and computer vision systems in order to identify the road, other traffic participants, and other relevant features or obstacles. The control algorithms on board these vehicles make use of explicit planning of a vehicle trajectory, such that feasibility and collisions can be checked prior to committing to the planned motion. This allows autonomous vehicles to handle a much wider class of traffic scenarios, compared to cooperative vehicles. However, in contrast to cooperative vehicles, the framework of autonomous vehicles is typically not aimed towards obtaining a string-stable traffic system of multiple autonomous vehicles. Additionally, these classes of vehicles use different automation frameworks: Cooperative vehicles in platoons make use of feedforward and feedback control, whereas autonomous vehicles make use of explicit trajectory planning. A promising way forward is to integrate the string stability objective and the communication aspect of the cooperative approach in the framework of autonomous vehicles. This would combine the versatility of the autonomous vehicle with the road capacity benefits of the cooperative vehicle. Such a vehicle would be capable of navigating the roads autonomously, and when encountering other equipped vehicles, decide to start platooning. Additionally, it can decide to overtake vehicles (e.g., heavy duty road vehicles) if these vehicles drive too slow to start a time-efficient platoon, or decide to break up a platoon if the preceding vehicle demonstrates behavior that is undesired for the host vehicle. The development of a unified framework for both cooperative and autonomous vehicles is therefore the main focus of the work presented here. This framework is designed to plan a trajectory relative to a reference path and reference velocity, which has been shown in literature to provide good results for autonomous vehicles. The trajectory planning is performed in a Frenet frame with respect to a nominal path, such that the vehicle will always follow the general direction of the road. B-splines are used to construct the planned trajectories, as these allow the trajectories to be efficiently communicated between vehicles since only a few parameters are required. This allows the required communication bandwidth to be low. The automated vehicles that utilize this framework simultaneously generate both cooperative as well as autonomous trajectories. This allows them to decide to break up the platoon at any moment if needed and instead utilize the autonomous trajectories. This vehicle could then potentially become, the lead vehicle of the newly formed platoon. The generation of the autonomous trajectory of the automated vehicle is the lowest cost trajectory of a set of potential candidate trajectories. This set also always includes trajectories that perform emergency braking or lateral deviations to prevent possible collisions. A cost function is used to select the most-comfortable, feasible and collision free trajectory. The cost functions for both cooperative and autonomous trajectories are designed in such a way that the selected trajectories in consecutive planning cycles are similar within the class of B-spline functions. This is referred to as temporal consistency. Temporal consistency is important as the selected trajectories are communicated to following vehicles, which in turn generate the cooperative trajectories based on the received information. For cooperative trajectories, timing is critical due to the small inter vehicle distance. The trajectories are constructed by means of a closed form computation, such that computation time can be guaranteed to satisfy the real time implementation requirements. Moreover, the communicated trajectories also include the time of construction on a common clock, which is obtained via the Global Positioning System (GPS). As a result, the cooperative vehicles are aware of the time delay between the planning of the preceding vehicle and the planning of the host vehicle, such that they can compensate for it. A challenge of adopting the trajectory planning framework for cooperative driving at short inter vehicle distances is the required computation time. The trajectories are typically updated 5 times per second, whereas distance information of the radar is typically updated around 20 times per second. To overcome the drawback of reduced control bandwidth, a method of time scaling is derived. This time scaling algorithm uses the measured inter vehicle distance to scale the time of the selected trajectory. In doing so, it allows the vehicle to slow down if needed, while maintaining the same geometric path as originally planned. Implementing the time scaling mechanism in the framework of the cooperative automated vehicle allows the use of the most up-to-date radar information available. To validate this theoretical framework, two full-scale prototype demonstrator platforms have been developed in the form of two modified Renault Twizy’s. The experiments performed with these demonstrator platforms show that the theoretical framework can be used for cooperative and automated driving at short inter-vehicle distances. Additionally, both the gap closing strategy as well as the time scaling mechanism are validated. The time scaling mechanism also showed promising results in overcoming system faults in the lead vehicle, where the lead vehicle would significantly deviate from its communicated trajectory. Summarizing, this thesis focusses on the development of a unified framework including both cooperative as well as autonomous vehicles, as well as the experimental validation thereof. The framework allows a single vehicle to act both as an autonomous vehicle, or demonstrate vehicle-following behavior at short inter vehicle distances, while considering string stability. The flexibility of decision making of the autonomous vehicle is retained, such that versatile traffic scenarios can be handled. Experimental validation demonstrates that the developed framework has great potential for further improvement for the next generation of automated vehicles.
Franz Lampel
added a research item
In orthogonal time frequency space (OTFS) modulation , information-carrying symbols reside in the delay-Doppler (DD) domain. By operating in the DD domain, an appealing property for communication arises: time-frequency (TF) disper-sive channels encountered in high mobility environments become time-invariant. The time-invariance of the channel in the DD domain enables efficient equalizers for time-frequency dispersive channels. In this paper, we propose an OTFS system based on the discrete Zak transform. The presented formulation not only allows an efficient implementation of OTFS but also simplifies the derivation and analysis of the input-output relation of TF dispersive channel in the DD domain.
Breno Beirigo
added a research item
Current mobility services cannot compete on equal terms with self-owned mobility products concerning service quality. Due to supply and demand imbalances, ridesharing users invariably experience delays, price surges, and rejections. Traditional approaches often fail to respond to demand fluctuations adequately since service levels are, to some extent, bounded by fleet size. With the emergence of autonomous vehicles (AVs), however, the characteristics of mobility services change, and new opportunities to overcome the prevailing limitations arise. This thesis proposes a series of learning- and optimization-based strategies to help autonomous transportation providers meet the service quality expectations of diversified user bases. We show how autonomous mobility-on-demand (AMoD) systems can develop to revolutionize urban transportation, improving reliability, efficiency, and accessibility.
Francesco Walker
added a research item
Trust predicts the disuse and misuse of automated vehicles. While a lack of trust may induce drivers to not use the automated vehicle's functionalities, excessive trust can lead to dangerous outcomes, with drivers using the system in ways that were not originally intended by its designers. This dissertation explores new ways through which trust can be reliably measured and aligned with the true capabilities of the automated driving system.
Francesco Walker
added a research item
To maximize road safety, driver trust in an automated vehicle should be aligned with the vehicle's technical reliability, avoiding under-and over-estimation of its capabilities. This is known as trust calibration. In the study reported here, we asked how far participant assessments of vehicle capabilities aligned with those of the engineers. This was done by asking the engineers to rate the reliability of the vehicle in a specific set of scenarios. We then carried out a driving simulator study using the same scenarios, and measured participant trust. The results suggest that user trust and engineer perceptions of vehicle reliability are often misaligned, with users sometimes under-trusting and sometimes over-trusting vehicle capabilities. On these bases, we formulated recommendations to mitigate under-and over-trust. Specific recommendations to improve trust calibration include the adoption of a more defensive driving style for first-time users, the visual representation of the objects detected by the automated driving system in its surroundings in the Human Machine Interface, and real-time feedback on the performance of the technology.
Breno Beirigo
added a research item
Residents of cities' most disadvantaged areas face significant barriers to key life activities, such as employment, education, and health-care, due to the lack of mobility options. Shared autonomous vehicles (SAVs) create an opportunity to overcome this problem. By learning user demand patterns, SAV providers can improve regional service levels by applying anticipatory relocation strategies that take into consideration when and where requests are more likely to appear. The nature of transportation demand, however, invariably creates learning biases towards servicing cities' most affluent and densely populated areas, where alternative mobility choices already abound. As a result, current disadvantaged regions may end up perpetually underserviced, therefore preventing all city residents from enjoying the benefits of autonomous mobility-on-demand (AMoD) systems equally. In this study, we propose an anticipatory rebalancing policy based on an approximate dynamic programming (ADP) formulation that processes historical demand data to estimate value functions of future system states iteratively. We investigate to which extent manipulating cost settings, in terms of subsidies and penalties, can adjust the demand patterns naturally incorporated into value functions to improve service levels of disadvantaged areas. We show for a case study in the city of Rotterdam, The Netherlands, that the proposed method can harness these cost schemes to better cater to users departing from these disadvantaged areas, substantially outperforming myopic and reactive benchmark policies.
Frederik Schulte
added 3 research items
Cooperation is important in order to find efficient vehicle routing solutions for the growing transportation market. Increasingly, platforms emerge as facilitators for this kind of collaborative transportation. However, individual actors connected to a platform might refuse to share (parts of) their information due to reasons of competition. Though the need for realistic information sharing models is widely acknowledged by transportation researchers and practitioners, the precise value of such information is mostly unknown. We investigate the quality of solutions that can be obtained when different types and levels of carrier data are available. We consider an auction-based Multi-Agent System to solve large-scale, dynamic pickup and delivery problems, and vary whether carriers' positions or route plans are available, and whether carriers are fully cooperative or more competitive in placing their bids by sharing or hiding their marginal costs. In total, we evaluate 9 different information sharing policies. The availability of vehicle position and route plan information turns out to be important for decreasing total route costs, and sharing marginal costs has a positive impact on service level. We provide detailed insights into trade-offs of carriers' confidentiality concerns and a range of system performance objectives (service level, travel costs, and carrier profits) under different circumstances (various numbers of auctions, penalties for rejected orders, emission, or congestion fines, and different problem characteristics). Based on these results, platform providers can stimulate sthe haring of certain information to improve the total system efficiency.
The transportation market requires collaboration to improve efficiency and reduce emissions. Individual carriers, however, might be hesitant to share their private information on a transportation platform. Although a few articles have investigated the value of information sharing, they assume identical behaviour of all carriers. In practice, nonetheless, some carriers are open to share more information than others. We consider such a hybrid information sharing setting and investigate the value of information sharing dependent on what other carriers are willing to share. We propose a Multi-Agent System in which carriers and customers interact and vary carriers’ willingness to share information about vehicle locations and marginal costs. The results show that sharing full route plans is always beneficial for individual carriers, independent of what position information other carriers share. However, to increase the total profit in scenarios with limited interaction, at least 50% of the carriers need to share full plans instead of only current positions. Furthermore, about 60% of the carriers need to be stimulated to share full cost information for solutions with maximal service level, although it might be unprofitable for themselves.
Frederik Schulte
added 2 research items
Current mobility services cannot compete on equal terms with self-owned mobility products concerning service quality. Due to supply and demand imbalances, ridesharing users invariably experience delays, price surges, and rejections. Traditional approaches often fail to respond to demand fluctuations adequately since service levels are, to some extent, bounded by fleet size. With the emergence of autonomous vehicles, however, the characteristics of mobility services change and new opportunities to overcome the prevailing limitations arise. In this paper, we consider an autonomous ridesharing problem in which idle vehicles are hired on-demand in order to meet the service level requirements of a heterogeneous user base. In the face of uncertain demand and idle vehicle supply, we propose a learning-based optimization approach that uses the dual variables of the underlying assignment problem to iteratively approximate the marginal value of vehicles at each time and location under different availability settings. These approximations are used in the objective function of the optimization problem to dispatch, rebalance, and occasionally hire idle third-party vehicles in a high-resolution transportation network of Manhattan. The results show that the proposed policy outperforms a reactive optimization approach in a variety of vehicle availability scenarios while hiring fewer vehicles. Moreover, we demonstrate that mobility services can offer strict service level contracts (SLCs) to different user groups featuring both delay and rejection penalties.
Franz Lampel
added 2 research items
This paper describes an FMCW based radar and communication (RadCom) system and addresses the challenges in the synchronization of multiple units for communication functionality. We proposed a novel technique to detect the FMCW RadCom signal at the communication receiver and derive the detection and false alarm probabilities of it. Moreover, to achieve fine synchronization between transmit and receive devices, a novel approach based on FMCW RadCom signal time of arrival estimation is proposed. The potential capability of a RadCom system is experimentally demonstrated for the first time by a set of automotive-grade mmWave radars with GPS-based synchronization.
The emerging trend of autonomously driving vehicles brings an increasing need for communication. In a setting comprising connected vehicles, information about the environment and vehicles themselves will be shared with other vehicles. Existing communication standards may not be able to meet this growing demand on communication bandwidth. Radar embedded communication can help overcome this bottleneck at the expense of degraded radar performance. In this work we introduce and demonstrate a novel radar signal processing technique to compensate for the self-interference due to the communication content on a chirp modulated radar signal. Recovery of the radar performance is demonstrated by simulations for the continuous phase modulation on the chirp modulation.
Francesco Walker
added a research item
One major concern with driving simulator studies is the lack of perceived risk for participants. This has led some authors to question the behavioural validity of simulator-based research. In this study, we investigated this concern by compensating for the possible perceived absence of risk with an anxiety-inducing risk factor: Participants were told that if they had a collision, they would receive a mild electric shock. We hypothesised that the addition of the new risk factor would increase participants' 'sense of presence' - the feeling of truly being and belonging in the virtual environment. We also analysed their driving behaviour, physiology, anxiety, and workload. Overall, we observed very few differences between the threat and the control group: Both reported a strong sense of presence. This suggests that, even without the risk of physical harm, mid-level driving simulators already elicit a strong sense of presence and that the 'lack of physical crash risk' is unlikely to affect study results.
Anika Boelhouwer
added 2 research items
As commercial cars start to include more automated functions it becomes difficult for drivers to understand how and when to use them safely. While general HMI recommendations for partially automated cars have been made, it is unclear how drivers should be supported during the initial use period. Recommendations for a tutor system that guides drivers in their initial use of partially automated cars are necessary. To gain inspiration for such a tutor system, we examined the existing communication loop of driving instructors and their students. Driving instructors and their students were video recorded during regular driving lessons. The tutoring patterns that were found (i.e. situation and student adaptive feedback, student adaptive tasks, body movements for correcting and requesting actions) during the initial qualitative analysis are discussed. Furthermore, we suggest methods how to implement the tutoring patterns in a tutor system to support drivers in the use partially automated cars.
Cars that include combinations of automated functions, such as Adaptive Cruise Control (ACC) and Lane Keeping (LK), are becoming more and more available to consumers, and higher levels of automation are under development. In the use of these systems, the role of the driver is changing. This new interaction between the driver and the vehicle may result in several human factors problems if not sufficiently supported. These issues include driver distraction, loss of situational awareness and high workload during mode transitions. A large conceptual gap exists on how we can create safe, efficient and fluent interactions between the car and driver both during automation and mode transitions. This study looks at different HMIs from a new perspective: Embodied Interaction. The results of this study identify design spaces that are currently underutilized and may contribute to safe and fluent driver support systems in partially automated cars.
Debargha Dey
added a research item
This paper discusses whether the knowledge of the driving mode of an approaching vehicle (manual vs. automated) influences pedestrians’ decisions while crossing a street. Additionally, the paper explores how different appearances and driving behaviours of vehicles interact with driving mode in affecting pedestrians’ road-crossing behaviours. In a video-based experiment with sixty participants, two vehicles with different appearances (a BMW 3 and a Renault Twizy) were presented as either manually-driven or automated vehicles. Both vehicles displayed either yielding or non-yielding behaviour on a straight road devoid of other traffic. Participants were asked to indicate whether they would cross the street in front of the approaching vehicle, at different distances ranging from 45 m to 1.5 m. The results showed that there was no significant influence of the knowledge of the driving mode (manually-driven vs automated) on pedestrians’ willingness to cross the street at any distance. The vehicle’s behaviour (whether it is maintaining speed or yielding) played a dominant role in pedestrians’ decision to cross a road, and this was similar for both modes and both vehicles, at all distances. However, results suggested that in situations and at distances when the intent of the vehicle was not fully clear by the behaviour of the car alone, there were differences between the two vehicles at certain distances, which could be attributed to the differences in their appearance such as size, aggressiveness and novelty. A futuristic-looking vehicle inspired less confidence in road-crossing situations compared to an ordinary-looking vehicle. Additionally, a novel and futuristic-looking vehicle appeared to make it easier for people to believe that it is an automated vehicle. We conclude by discussing design implications for the development of external HMIs automated vehicles.
Robbin van Hoek
added a research item
Automated vehicles are used to improve traffic throughput and increase safety on roads. These automated vehicles should navigate both on highways and urban roads, and adopt motion planners to ensure collision free and comfortable trajectories. This work extends a general planning framework for an autonomous vehicle with the communication aspect of cooperative vehicles, to improve the behavior of a string of automated vehicles. By including vehicle to vehicle communication, the autonomous vehicle is better suited for vehicle following with small inter-vehicle gaps. Considerations regarding the behavior of the multi-vehicle platoon are presented, supported with numerical simulations.
Francesco Walker
added 2 research items
Studies show that drivers' intention to use automated vehicles is strongly modulated by trust. It follows that their benefits are unlikely to be achieved if users do not trust them. To date, most studies of trust in automated vehicles have relied on self-reports. However, questionnaires cannot capture real-time changes in drivers' trust, and are hard to use in applied settings. In previous work, we found evidence that gaze behaviour could provide an effective measure of trust. In this study we tested whether combining gaze behaviour with Electrodermal Activity could provide a stronger metric. The results indicated a strong relationship between self-reported trust, monitoring behaviour and Electrodermal Activity: The higher participants' self-reported trust, the less they monitored the road, the more attention they paid to a non-driving related secondary task, and the lower their Electrodermal Activity. We also found evidence that combined measures of gaze behaviour and Electrodermal Activity predict self-reported trust better than either of these measures on its own. These findings suggest that such combined measures have the potential to provide a reliable and objective real-time indicator of driver trust.
Breno Beirigo
added 3 research items
In the realm of human urban transportation, many recent studies have shown that comparatively smaller fleets of shared autonomous vehicles (SAVs) are able to provide efficient door-to-door transportation services for city dwellers. However, because of the steady growth of e-commerce and same-day delivery services, new city logistics approaches will also be required to deal with last-mile parcel delivery challenges. We focus on modeling a variation of the people and freight integrated transportation problem (PFIT problem) in which both passenger and parcel requests are pooled in mixed-purpose compartmentalized SAVs. Such vehicles are supposed to combine freight and passenger overlapping journeys on the shared mobility infrastructure network. We formally address the problem as the share-a-ride with parcel lockers problem (SARPLP), implement a mixed-integer linear programming (MILP) formulation, and compare the performance of single-purpose and mixed-purpose fleets on 216 transportation scenarios. For 149 scenarios where the solver gaps of the experimental results are negligible (less than 1%), we have shown that mixed-purpose fleets perform in average 11% better than single-purpose fleets. Additionally, the results indicate that the busier is the logistical scenario the better is the performance of the mixed-purpose fleet setting.
Autonomous vehicles (AVs) are expected to widely re-define mobility in the future, transforming many solutions into autonomous services. Nonetheless, this development requires an expected transition phase of several decades in which some regions will provide sufficient infrastructure for AV movements, while others will not support AVs yet. In this work, we propose an operational planning model for mobility services operating in regions with AV-ready and not AV-ready zones. To this end, we model detailed automated driving areas and consider a heterogeneous fleet comprised of three vehicle types: autonomous, conventional, and dual-mode. While autonomous and conventional vehicles can only operate in their designated areas, dual-mode vehicles service zone-crossing demands in which both human and autonomous driving are required. For such a hybrid network, we introduce a new mathematical planning model based on a site-dependent variant of the heterogeneous dial-a-ride problem (HDARP). With a numerical study for the city of Delft, The Netherlands, we provide insights into how operational costs, service levels, and fleet utilization develop under 405 scenarios of multiple infrastructural settings and technology costs.
With the popularization of transportation network companies (TNCs) (e.g., Uber, Lyft) and the rise of autonomous vehicles (AVs), even major car manufacturers are increasingly considering themselves as autonomous mobility-on-demand (AMoD) providers rather than individual vehicle sellers. However, matching the convenience of owning a vehicle requires providing consistent service quality, taking into account individual expectations. Typically, different classes of users have different service quality (SQ) expectations in terms of responsiveness, reliability, and privacy. Nonetheless, AMoD systems presented in the literature do not enable active control of service quality in the short term, especially in light of unusual demand patterns, sometimes allowing extensive delays and user rejections. This study proposes a method to control the daily operations of an AMoD system that uses the SQ expectations of heterogeneous user classes to dynamically distribute service quality among riders. Additionally, we consider an elastic vehicle supply, that is, privately-owned freelance AVs (FAVs) can be hired on short notice to help providers meeting user service-level expectations. We formalize the problem as the dial-a-ride problem with service quality contracts (DARP-SQC) and propose a multi-objective matheuristic to address real-world requests from Manhattan, New York City. Applying the proposed service-level constraints, we improve user satisfaction (in terms of reached service-level expectations) by 53% on average compared to conventional ridesharing systems, even without hiring additional vehicles. By deploying service-quality-oriented on-demand hiring, our hierarchical optimization approach allows providers to adequately cater to each segment of the customer base without necessarily owning large fleets.
Anika Boelhouwer
added 2 research items
Partially automated car systems are expected to soon become available to the public. However, in order for any of the potential benefits of automated driving to arise, the driver and car need to establish effective, efficient and satisfactory interactions. Otherwise, the driver may rely too much on the automated car system, leading to dangerous situations or not relying on the system at all, making the automation pointless. This study studied whether the current method of providing information on (automated) car systems to drivers, which is mainly through owner’s manuals, can bring the driver’s mental model in accordance with the car’s capabilities. A total of 28 participants took part in a video- based driving simulator experiment. The participants were split into two groups: the first received no information about the system while the second did receive specific information about functionalities and system limitations. Each participant was seated in a driving simulator and experienced a partially automated car driving in city situations by means of videos projected on the outer screen. Participants were asked to indicate through the push of a button on the steering wheel if they felt that the car could no longer cope with the situation, and would take back control from the car if they were driving it on the real road. Each video was categorized as ‘requires a take-over’ or ‘does not require a take-over’ before the experiment, based on the system descriptions the participants received. Overall, the system information did not appear to support the participants in correctly deciding whether to take over or to rely on the system. The mental models of the participants did not seem to (sufficiently) change through the system information. Owner’s manuals may not be sufficient for future systems to provide drivers the necessary tools to be able to decide whether it is necessary to take back control of the car. In-vehicle support, tuned to the driver and the specific situation may be needed to safely guide this process.
Overtrust and undertrust are major issues with partially automated vehicles. Ideally, trust should be calibrated ensuring that drivers' subjective feelings of safety match the objective reliability of the vehicle. In the present study, we examined if drivers' trust toward Level 2 cars changed after on-road experience. Drivers' self-reported trust was assessed three times: before having experience with these vehicles, immediately after driving two types of vehicles, and two weeks after the driving experience. Analysis of the results showed major changes in trust scores after the on-road driving experience. Before experiencing the vehicles, participants tended to overestimate the vehicle capabilities. Afterwards they had a better understanding of vehicles' limitations, resulting in better calibrated trust.
Breno Beirigo
added a project goal
Address current transportation challenges regarding throughput and safety with an integrated approach to automated and cooperative driving. In i-CAVE, a Cooperative Dual Mode Automated Transport (C-DMAT) system is researched and designed, consisting of dual mode vehicles which can be driven automatically and manually to allow maximum flexibility. The program integrates technological road maps for automated and cooperative driving, accelerating the development of novel transportation systems addressing today and future mobility demands.
Access https://i-cave.nl/ for more information.