1. Industry Panel: Connected, cooperative and automated transport
Recent developments in telecommunication, sensor and information technologies have enabled substantial progress in the domain of automation of transport. Cooperative driving is technologically achievable today, but is only in a very early stage of deployment. Automated driving is on the horizon, and will still need substantial and longer-term development and testing to make even the high automation levels a reality in complex situations such as in urban environments, and in a transit period of only partial market penetration. Cooperative and automated transport are certainly complementary. They are expected to bring substantial benefits in terms of safety, comfort and (traffic and fuel) efficiency. Many challenges exist in this important domain. The workshop targets competing communication technologies, e.g. peer to peer (IEEE 802.11p), cellular network, and future 5G. The challenges of ITS applications towards automated driving (especially related to telecommunication) will be highlighted. Industry requirements will be analysed. 5G development status, especially from the perspective of automotive applications, will be updated. Innovation trends and topics for R&D; cooperation will be addressed. The workshop is expected to be very interactive. Participants will have an excellent opportunity to discuss with, and to challenge distinguished speakers from industry.
2. Human Factors in Intelligent Vehicles
The Workshop on Human Factors in Intelligent Vehicles (HFIV ’17) aims to foster discussion on issues related to the analysis of human factors in the design and evaluation of intelligent vehicles (IV) technologies, in a wide spectrum of applications and in different dimensions. It is expected to build upon a proper environment to disseminate knowledge and motivate interactions among the technical and scientific communities, practitioners and students, allowing state-of-the-art concepts and advances to be further developed and enhanced.
3. Eco-drive at Intersections with Connected,Cooperative and Automated Vehicle Technology
Today’s road traffic systems are facing challenges in cutting down congestion and contributing to environmental sustainability. Especially the urban environment is suffering from congested traffic and an inherent high level of emissions. Eco—driving systems aim at assisting drivers for vehicle operations can help improve fuel economy and consequently reduce emissions. The potential of eco-drive strategies for coping with these issues can be enlarged if combined with cooperative and connected systems based on communication technologies. Along these opportunities come a few challenges for authorities, industry, as well as scientific community. In terms of system design and control, current eco-drive systems need to be refined or even redesigned to better function under uncertainties in demand and mixed traffic conditions and to better cooperate with traffic signal control systems. From the performance assessment perspective, traffic flow models and simulation tools have been widely used to verify the performance of eco-drive systems, in particular taking into account the increasing trends in vehicle connectivity and automation. However, the validity of these models needs to be re-examined against field tests.This workshop focuses on eco-drive systems using connected, cooperative and automated vehicle technology, applied in a road network with intersections. The main goal is to share the state of the art in design, models, algorithms, simulation and field test of eco-drive systems, identify challenges and research needs, and encourage cross-disciplinary cooperation.
4. Workshop on Modelling, Analysis and Control of Intelligent Mobility-on-Demand Systems
After decades of little innovation, personal urban mobility is undergoing rapid transformations due to the introduction of disruptive technologies (e.g. connected and driverless cars), new IT applications (e.g. appbased services) but also due to changes in individual preferences and social behaviours, with a growing trend towards a shifting from car ownership to sharing. This gave new life to several mobility on demand (MOD) services which were ideated decades ago but never established themselves as viable mobility solutions and created new variations of them, such as ride-sharing, bike-sharing programs, car-pooling and car-sharing services, on-demand bus and delivery services, etc. The rapid growth and the forecasted (large) scale of these new mobility services is expected to radically change individual travel patterns, and conventional frameworks for the modelling, analysis, simulation and control of transportation systems are not appropriate any more. For instance, novel demand modelling tools are needed for measuring, modelling and predicting behavioural choice and individual preferences for the new mobility solutions, as well as forecasting the level of market uptake of the different mobility services. Similarly, new analytical models and simulation frameworks are required to accurately characterise the peculiar properties of MOD systems. Then, the insights obtained may serve as basic input to advanced optimization frameworks, which can provide decision tools for the planning and optimal operation of such systems. Key issues to address are infrastructure planning, fleet sizing and management, supply rebalancing, and efficient cooperation with other transportation modes (e.g. public transport). The goal of this workshop session is to provide a forum to exchange ideas, discuss solutions, and share experiences from industry, researchers and the public sector. We solicit original papers covering different aspects of MOD systems.
5. Deep Learning for Autonomous Driving
Deep learning has been progressing rapidly and disseminated via conferences like NIPS and CVPR. This workshop is an attempt to bridge the gap between latest research in Deep Learning and application to autonomous driving which is an active area of research in both academia and industry. The first success of Deep Learning was mainly in visual perception via CNNs which has enabled applications like semantic segmentation which wasn’t deemed possible before and expanding into classical geometric vision problems like Optical Flow and Structure from Motion. The other application areas like motion planning, sensor fusion, etc are in early stages of research. There is also the ambitious side of solving autonomous driving by a single deep learning model (end-to-end learning) and its variant of modular end-to-end with auxiliary losses for semantics. From a deployment perspective, processing power is still a bottleneck and there is steady increase of computational power where next generation platforms are targeting 10-100 TOPS.