remote driving
Recently Published Documents

Despite the progress in the development of automated vehicles in the last decade, reaching the level of reliability required at large-scale deployment at an economical price and combined with safety requirements is still a long road ahead. In certain use cases, such as automated shuttles and taxis, where there is no longer even a steering wheel and pedals required, remote driving could be implemented to bridge this gap; a remote operator can take control of the vehicle in situations where it is too difficult for an automated system to determine the next actions. In logistics, it could even be implemented to solve already more pressing issues such as shortage of truck drivers, by providing more flexible working conditions and less standstill time of the truck. An important aspect of remote driving is the connection between the remote station and the vehicle. With the current roll-out of 5G mobile technology in many countries throughout the world, the implementation of remote driving comes closer to large-scale deployment. 5G could be a potential game-changer in the deployment of this technology. In this work, we examine the remote driving application and network-level performance of remote driving on a recently deployed sub-6-GHz commercial 5G stand-alone (SA) mobile network. It evaluates the influence of the 5G architecture, such as mobile edge computing (MEC) integration, local breakout, and latency on the application performance of remote driving. We describe the design, development (based on Hardware-in-the-Loop simulations), and performance evaluation of a remote driving solution, tested on both 5G and 4G mobile SA networks using two different vehicles and two different remote stations. Two test cases have been defined to evaluate the application and network performance and are evaluated based on position accuracy, relative reaction times, and distance perception. Results show the performance of the network to be sufficient for remote driving applications at relatively low speeds (<40 km/h). Network latencies compared with 4G have dropped to half. A strong correlation between latency and remote driving performance is not clearly seen and requires further evaluation taking into account the influence of the user interface.

Even entirely driverless vehicles will sometimes require remote human intervention. Existing SA frameworks do not acknowledge the significant human factors challenges unique to a driver in charge of a vehicle that they are not physically occupying. Remote operators will have to build up a mental model of the remote environment facilitated by monitor view and video feed. We took a novel approach to “freeze and probe” techniques to measure SA, employing a qualitative verbal elicitation task to uncover what people “see” in a remote scene when they are not constrained by rigid questioning. Participants (n = 10) watched eight videos of driving scenes randomized and counterbalanced across four road types (motorway, rural, residential and A road). Participants recorded spoken descriptions when each video stopped, detailing what was happening (SA Comprehension) and what could happen next (SA Prediction). Participant transcripts provided a rich catalog of verbal data reflecting clear interactions between different SA levels. This suggests that acquiring SA in remote scenes is a flexible and fluctuating process of combining comprehension and prediction globally rather than serially, in contrast to what has sometimes been implied by previous SA methodologies (Jones and Endsley, 1996; Endsley, 2000, 2017b). Inductive thematic analysis was used to categorize participants’ responses into a taxonomy aimed at capturing the key elements of people’s reported SA for videos of driving situations. We suggest that existing theories of SA need to be more sensitively applied to remote driving contexts such as remote operators of autonomous vehicles.

These days, autonomous vehicles (AVs) technology has been improved dramatically. However, even though the AVs require no human intervention in most situations, AVs may fail in certain situations. In such cases, it is desirable that humans can operate the vehicle manually to recover from a failure situation through remote driving. Furthermore, we believe that remote driving can enhance the current transportation system in various ways. In this paper, we consider a revolutionary transportation platform, where all the vehicles in an area are controlled by some remote controllers or drivers so that transportation can be performed in a more efficient way. For example, road capacity can be effectively utilized and fuel efficiency can be increased by centralized remote control. However, one of the biggest challenges in such remote driving is the communication latency between the remote driver and the vehicle. Thus, selecting appropriate locations of the remote drivers is very important to avoid any type of safety problem that might happen due to large communication latency. Furthermore, the selection should reflect the traffic situation created by multiple vehicles in an area. To tackle these challenges, in this paper, we propose several algorithms that select remote drivers’ locations for a given transportation schedules of multiple vehicles. We consider two objectives in this system and evaluate the performance of the proposed algorithms through simulations. The results show that the proposed algorithms perform better than some baseline algorithms.

Developments in sensor technology, artificial intelligence, and network technologies like 5G has made remote operation a valuable method of controlling various types of machinery. The benefits of remote operations come with an opportunity to access hazardous environments. The major limitation of remote operation is the lack of proper sensory feedback from the machine, which in turn negatively affects situational awareness and, consequently, may risk remote operations. This article explores how to improve situational awareness via multimodal feedback (visual, auditory, and haptic) and studies how it can be utilized to communicate warnings to remote operators. To reach our goals, we conducted a controlled, within-subjects experiment in eight conditions with twenty-four participants on a simulated remote driving system. Additionally, we gathered further insights with a UX questionnaire and semi-structured interviews. Gathered data showed that the use of multimodal feedback positively affected situational awareness when driving remotely. Our findings indicate that the combination of added haptic and visual feedback was considered the best feedback combination to communicate the slipperiness of the road. We also found that the feeling of presence is an important aspect of remote driving tasks, and a requested one, especially by those with more experience in operating real heavy machinery.

Even entirely driverless vehicles will sometimes require remote human intervention. Existing SA frameworks do not acknowledge the significant human factors challenges unique to a driver in charge of a vehicle that they are not physically occupying. Remote operators will have to build up a mental model of the remote environment facilitated by monitor view and video feed. We took a novel approach to 'freeze and probe' techniques to measure SA, employing a qualitative verbal elicitation task to uncover what people ‘see’ in a remote scene when they are not constrained by rigid questioning. Participants (n=10) watched eight videos of driving scenes randomised and counterbalanced across four road types (motorway, rural, residential and A road). Participants recorded spoken descriptions when each video stopped, detailing what was happening (comprehension) and what could happen next (prediction). Participant transcripts provided a rich catalogue of verbal data reflecting clear interactions between different SA levels. This suggests that acquiring SA in remote scenes is a flexible and fluctuating process of combining comprehension and prediction globally rather than serially, in contrast to what has sometimes been implied by previous SA methodologies (Endsley, 2000; Endsley, 2017; Jones & Endsley, 1996). Participants’ responses were also categorised to form a ‘Taxonomy of SA’ aimed at capturing the key elements of people’s reported SA for videos of driving situations. We suggest that existing theories of SA need to be more sensitively applied to remote driving contexts such as remote operators of autonomous vehicles.