Source-linked AI summary
A Review of Sensing and Communication, Human Factors, and Controller Aspects for Information-Aware Connected and Automated Vehicles
Ankur Sarker, Haiying Shen, Mizanur Rahman, Mashrur Chowdhury, Kakan Dey, Fangjian Li, Yue Wang, Husnu S. Narman
TL;DR
The paper reviews how sensing and connectivity, human factors, and information-aware control relate to safe and efficient CAV operation. It synthesizes technologies, design considerations, challenges, and future directions, concluding that V2X information can enhance controller performance and that user comfort matters to CAV design.
Problem
Method
The authors review CAV sensing and communication technologies, human factors, and information-aware controller designs and criteria.
Source: §1 · pp. 2–3Results
The review concludes that V2X communication can enhance CAV controller performance by supplying more information, while V2I can support centralized control and global optimization.
Source: §5.1.3 · p. 20Takeaways & Limitations
The paper argues that combining connectivity-derived information with onboard sensing and user-comfort considerations can support safer, more efficient, and more comfortable CAV operations.
Source: §2 · p. 3
Abstract
from arXiv · showhide
Information-aware connected and automated vehicles (CAVs) have drawn great attention in recent years due to its potentially significant positive impacts on roadway safety and operational efficiency. In this paper, we conduct an in-depth review of three basic and key interrelated aspects of a CAV: sensing and communication technologies, human factors, and information-aware controller design. First, different vehicular sensing and communication technologies and their protocol stacks, to provide reliable information to the information-aware CAV controller, are thoroughly discussed. Diverse human factor issues, such as user comfort, preferences, and reliability, to design the CAV systems for mass adaptation are also discussed. Then, different layers of a CAV controller (route planning, driving mode execution, and driving model selection) considering human factors and information through connectivity are reviewed. In addition, critical challenges for the sensing and communication technologies, human factors, and information-aware controller are identified to support the design of a safe and efficient CAV system while considering user acceptance and comfort. Finally, promising future research directions of these three aspects are discussed to overcome existing challenges to realize a safe and operationally efficient CAV.
1 Introduction
The introduction frames safe and efficient fully automated vehicles as dependent on sensing and communication, human factors, and information-aware control. It reviews these interrelated areas, emphasizing user acceptance and the need to account for information from connectivity.
1 Introduction
- The introduction traces AV development from PATH’s six-vehicle platooning application in 1986 to later prototypes designed for mixed traffic [1, 2, 3].It also notes that sensor failure remains a risk, while computation, communication, and backup sensors may reduce its probability.Source: §1 · p. 2
- More than 75% of US drivers were reported uncomfortable using any kind of AV, while recent crashes highlight the consequences of uncertainty in traffic environments.The paper argues that understanding sensing and communication, human factors, and information-aware control is necessary for fully automated vehicles.Source: §1 · p. 2
- CAV adoption depends on user comfort, trust, and preferences, which vary with individual and societal characteristics and shape acceptable vehicle dynamics and headways.The introduction also calls for driver behavior models, including car-following and lane-changing, in path planning.Source: §1 · p. 2
- V2X communication lets controllers use information beyond onboard sensing, and the paper reviews control schemes that use external information across controller layers.The introduction identifies sensing and communication, human acceptance, and controller design as areas for further advancement.Source: §1 · p. 2
- The paper reviews sensing and communication technologies, human factors, and controller-design criteria for fully automated vehicles, complementing earlier reviews of ITS and CACC.The authors distinguish their focus from their prior CACC survey.Source: §1 · pp. 2–3
2 Focus of the Paper
The paper presents CAV operation as an integration of onboard sensing, wireless connectivity, and user-aware control. A roadway-incident scenario illustrates how these elements can inform safe and comfortable vehicle actions.
2 Focus of the Paper
- Onboard sensors have line-of-sight and range limits, while V2X can provide a broader view and roadway information such as signal timing, incidents, and queues.The paper describes the resulting information-aware controller as supporting safer, more efficient, and more comfortable operations than a conventional AV controller.Source: §2 · p. 3
- In the incident scenario, an RSU sends a warning to a platoon leader over V2I, which relays it over V2V before the platoon slows and changes lanes.The controller uses incident information and user preferences to calculate trajectories that satisfy comfort and safety requirements.Source: §2 · p. 4
- The review examines the current status, challenges, and future directions of sensing and communication, human factors, and information-aware AV controllers.The scenario highlights the need for reliable, low-latency communication and acceptable vehicle dynamics and string stability under user preferences.Source: §2 · p. 4
3 Sensor and Communication Technologies
This section surveys sensing and communication technologies for CAV safety and traffic efficiency. It covers sensor and communication research, challenges, and future directions.
3 Sensor and Communication Technologies
- A USDOT breakdown analysis of 2004–2008 crash data states that in-vehicle communication and sensing technologies could help avoid up to 79% of all traffic accidents.The cited applications include forward-collision, emergency-braking, blind-spot, left-turn, and lane-departure warnings.Source: §3 · p. 4
- Sensor and communication technologies can also harmonize vehicle speeds, control traffic signals, and provide collision warnings and improved riding experiences.Source: §3 · p. 4
- The section reviews sensing technologies and V2V and V2I communications, including their research trends, challenges, and future directions.It also discusses promising technologies for future-generation CAV systems.Source: §3 · p. 4
I. Sensor Technologies
The section surveys onboard sensors and the challenges that shape their use in CAVs. It highlights sensor trade-offs, environmental constraints, and research needs in fusion, security, deployment, and cost.
I. Sensor Technologies
- Onboard radar, cameras, and LIDAR monitor vehicle surroundings and support collision-avoidance warnings that let drivers respond to hazards.Source: §I · p. 4
- Short-range radar detects objects within 20 meters, while long-range radar reaches 150 meters; LIDAR perception ranges from 10 to 200 meters.The cited radar uses include parking assistance and blind-spot warning for short-range systems and forward-collision warning and intersection management for long-range systems.Source: §I · pp. 4–5
- Cameras provide color and spatial information for lane and sign detection, with background-subtraction, feature-based, and motion-based approaches used to detect moving objects.Camera configurations and placement vary by purpose, including stereo cameras for a wider view and infrared cameras for night or bad weather.Source: §I · pp. 4–5
- Ultrasonic acoustic sensors measure object distance from sound-wave echoes and are used in parking, backing, lane maintenance, and cruise control.Cited work also uses acoustic sensing to extract spatial features and estimate road congestion and vehicle speed distribution.Source: §I · p. 5
- Sensor selection involves trade-offs: active sensors can suffer interference, while camera sensors need more computation and are sensitive to lighting.The paper says selection should account for dynamic range, spatial resolution, spectral sensitivity, and computational capabilities.Source: §I · p. 5
- In-vehicle sensors can be obstructed by nearby vehicles, and camera performance can be affected by lighting and weather conditions.The section also notes poor performance by in-vehicle sensors in heavy rain, fog, and snow, and reduced camera performance from shadows and light transitions.Source: §I · p. 5
- The authors call for further sensor-fusion research, citing improved classification and robustness from combining sensors.Source: §I · p. 6
- Future work includes sensor-network security, real-traffic verification and benchmarks, and smaller, more affordable sensor hardware.The paper notes that a rich sensor set can increase hardware costs and support needs.Source: §I · p. 6
II. Communication Technologies
The section reviews vehicular communication architectures and protocols, from DSRC-based V2X links to newer and heterogeneous networks. It also surveys congestion and connectivity challenges and identifies research priorities for dependable communication in automated traffic.
Communication architecture
- DSRC-based communication uses IEEE 802.11p and IEEE 1609 layers to form the WAVE suite for secure V2V and V2I services.The section describes DSRC channels and the roles of on-board and roadside units in vehicle networking.Source: §II · pp. 6–7
Vehicle-to-vehicle communication
- V2V beacon schemes use contention-free TDMA or contention-based channel adjustments, while congestion-control studies include proactive and feedback-based methods.Changing traffic density and unreliable channels make effective periodic beacon transmission an open problem.Source: §II · pp. 7–8
Vehicle-to-infrastructure and extended networks
- RSUs can relay messages and extend V2V coverage, while hybrid DSRC, LTE, and Wi-Fi networks and vehicle-cloud models offer additional connectivity.Vehicle-cloud communication can provide services but has higher transmission delay and cost than V2V or V2I links.Source: §II · pp. 7–9
Next-generation communication technologies
- The review discusses Li-Fi, LTE Advanced Pro, and 5G as prospective technologies for vehicular communication, including low-latency and high-reliability scenarios.The passage reports a 5G estimated round-trip latency below 1 ms and identifies URLLC for high-mobility scenarios.Source: §II · pp. 9–10
Challenges
- High mobility, channel congestion, and cellular-network latency challenge reliable communication, particularly for safety messages and dense traffic.The review notes frequent disconnections, packet loss and access delay, and bottlenecks in cellular downlink and uplink channels.Source: §II · p. 10
Future research directions
- Future work should develop robust heterogeneous links, validate channel-allocation schemes in real traffic, and strengthen security and privacy for connected vehicles.The review also calls for context-aware resource scheduling and scalable data dissemination.Source: §II · pp. 10–11
4 Human Factors
Human behavior remains central to CAV design because driver errors and technology-related distractions coexist with complex interactions between people and vehicles.
Human factors
- Driver errors contribute to traffic incidents, while entertainment, handheld devices, and navigation systems introduce additional forms of distraction.The passage identifies long reaction times, distracted driving, and driving under the influence as driver limitations.Source: §4 · p. 11
- Understanding driver behavior and its interaction with the vehicle is a recognized challenge in both traditional and automated vehicle design.Source: §4 · p. 11
- CAV design therefore needs to account for human behavior and provide reliable human-machine interfaces.Source: §4 · p. 11
4.1 Current Status
The section organizes CAV human factors around system design for user expectations and user adaptation to the designed system. It reviews comfort, trust, behavior, and situational awareness as considerations for acceptance and safe interaction.
4.1 Current Status
- Human factors are grouped into designing for user expectations and helping users adapt, with comfort, trust, behavioral adaptation, and situational awareness among the considerations.The review identifies acceptance, workload, usability, transitions, and responses to failures as human-factor issues.Source: §4.1 · pp. 11–12
4.1.1 Design of a CAV System Considering User Expectation
- CAV longitudinal and lateral controls should reflect user comfort and preferences, while car-following and lane-changing models represent driving behavior across conditions.The reviewed car-following models differ in features; one field evaluation found the IDM more comfortable than two other controllers, but slower and with larger CACC headways.Source: §4.1.1 · pp. 12–13
- Trust is associated with system accuracy and reliability, and continuous information about vehicle state may be more effective than immediate failure warnings.The review lists route accuracy, roadway information, failure detection, and user familiarity among factors related to desirability and reliability.Source: §4.1.1 · pp. 13–14
4.1.2 Adaption to the Designed CAV System by User
- HMI supports users shifting between actuator and supervisory roles, while regular information can help maintain awareness of surroundings and vehicle state.The review describes situational awareness as perception, comprehension, and prediction; engagement shifts, non-driving tasks, and mode confusion are concerns during emergencies.Source: §4.1.2 · p. 14
4.2 Challenges
The review identifies challenges in modeling preference-sensitive driving, preserving user awareness during failures, and designing reliable human-machine interaction.
4.2.1 User Preferences Modelling
- Preference-sensitive longitudinal and lateral control must detect nearby conditions and predict other vehicles’ intentions, especially during merging and lane changes.Failures in detection or prediction can destabilize vehicle dynamics; conservative responses to uncertainty may reduce lateral efficiency, comfort, and acceptance.Source: §4.2.1 · pp. 14–15
4.2.2 CAV’s System Failure
- During system failures, CAVs need suitable methods to support user situational awareness and use V2V or V2I information for early unsafe-situation warnings.Source: §4.2.2 · p. 15
4.2.3 HMI
- HMI design needs repeatable results and further study of user errors, over- or under-reliance, feedback interfaces, and human-centered control.Source: §4.2.3 · p. 15
4.3 Future Research Directions
The review identifies future directions aimed at reliable, user-acceptable CAV automation: model individual preferences, prepare users to respond to system failures, and integrate assistance systems in the HMI.
4.3.1 CAV User Preferences Modelling using Artificial Intelligence
- Future work should model speed, acceleration, gap, and lane-change preferences using human-centered learning and trajectory data.Preference-aware merging and diverging models should account for gap acceptance and available target-lane gaps.Source: §4.3.1 · p. 15
4.3.2 Engaging in a Case of CAV’s System Failure
- For safety-critical failures, engagement monitoring and visual, audible, or vibration warnings can help disengaged users regain awareness and control.The text notes that users’ ability to resume supervision is limited by their capacity to stay alert while disengaged.Source: §4.3.2 · p. 15
4.3.3 Integration of Multiple Assistance Systems in the HMI
- Future HMI work should integrate multiple assistance systems while providing reliable operation and accurate information, including about malfunctions.Source: §4.3.3 · p. 15
5 Information-aware Controller Design
The controller must issue commands despite sensing-related uncertainty, while V2X can supply information beyond onboard sensors. The review examines how wireless communication is used, current challenges, and future directions.
5 Information-aware Controller Design
- Sensing-device limitations create environmental uncertainty that can compromise sensor-based AV efficiency when driving safety is prioritized.Source: §5 · p. 16
- V2X communication provides information beyond onboard sensing, which the information-aware controller can use to improve driving performance.Source: §5 · p. 16
- The section reviews current V2X information use, controller challenges, and future research directions.Source: §5 · p. 16
5.1 Current Status
The review describes information-aware control across route planning, driving-mode selection, and execution, and concludes that V2X can improve controller performance under traffic uncertainty; V2I can also support centralized global optimization.
5.1 Current Status
- The five-layer controller uses perception and localization to gather and place information, then plans routes, selects driving modes, and executes them.The review focuses on information-aware design in the latter three layers.Source: §5.1 · p. 16
5.1.1 Route Planning Layer
- V2X route planning can use real-time traffic information for individual travel-time optimization or centralized route assignment that accounts for system-wide traffic effects.Reviewed approaches include predicted segment travel times, urgency-based route assignment, congestion-aware costs, and route replanning constrained by RSU coverage.Source: §5.1.1 · pp. 16–17
5.1.2 Driving Mode Selection Layer
- V2V trajectory and motion-state sharing can help driving-mode selection resolve uncertainty about neighboring vehicles and support speed adjustment or lane changing.Shared planned trajectories can serve as references for collision-free trajectories and help resolve whether a slower vehicle will change lanes.Source: §5.1.2 · pp. 17–18
- V2I-based centralized traffic controllers can optimize traffic flow using segment speed, lane-changing flow, and ramp-entry rate, including in mixed traffic.Source: §5.1.2 · p. 18
5.1.3 Driving Mode Execution Layer
- In vehicle following, CACC uses V2X information and feedforward control to support shorter time headways with string stability than ACC.Reviewed CACC work also uses reinforcement learning and MPC; MPC objectives can include jerk minimization for occupant comfort.Source: §5.1.3 · pp. 18–19
- For lane changing, V2V slot negotiation can avoid conservative assumptions about neighbors’ maximum acceleration and give target-lane vehicles more preparation time, reducing velocity fluctuation.Proactive merging can also use RSU instructions to coordinate vehicle speeds and leave a gap for the merging vehicle.Source: §5.1.3 · p. 19
- At intersections, SPaT-informed speed choices and V2V or V2I coordination support execution strategies for signalized and unsignalized settings.The reviewed centralized intersection controller manages non-overlapping trajectories subject to time-headway and acceleration constraints.Source: §5.1.3 · pp. 19–20
5.1 Current Status
- V2X can enhance controller efficiency and performance under roadway uncertainty, while V2I can support centralized control that optimizes overall traffic conditions.Source: §5.1.3 · p. 20
5.2 Challenges
The review identifies three controller challenges: balancing centralized and distributed control, addressing communication imperfections, and accommodating mixed traffic with non-connected vehicles.
5.2.1 Centralized versus Distributed Controller
- Centralized control may optimize traffic flow, but liability, bandwidth, and computation constraints complicate its use relative to distributed control.The review also notes that individual vehicles may not process all communication information in real time.Source: §5.2.1 · p. 20
5.2.2 Communication Imperfection
- Communication delay, shadow fading, and interference can degrade vehicle-following safety, efficiency, and comfort, while effects on route planning and mode selection remain understudied.The reviewed work notes that many studies assume perfect communication, which the authors characterize as unrealistic.Source: §5.2.2 · p. 20
5.2.3 Mixed Traffic Scenario
- Non-connected vehicles limit centralized state sharing and can force V2V cooperative controllers to fall back to sensor-based control, potentially compromising roadway efficiency.The authors argue that CAV and non-connected vehicle coexistence should be considered as CAVs enter the market gradually.Source: §5.2.3 · p. 20
5.3 Future Research Directions
Although current automated-car controllers meet various objectives, the review identifies future work in hybrid V2V/V2I control, resilience to imperfect communication, and accounting for non-connected drivers.
5.3.1 Hybrid Structure with Centralized and Distributed Control Scheme
- A unified hybrid V2V/V2I control structure could combine centralized computation and richer environmental information with distributed control, and may help address cybersecurity challenges.The centralized controller is located at a traffic management center; CAVs may also have high-computation on-board units.Source: §5.3.1 · p. 21
5.3.2 Communication Imperfection Resilient Control Scheme
- Communication loss remains possible despite 5G bandwidth and low latency; resilient control beyond longitudinal regulation and urgency-based message scheduling remain future directions.The review notes that cyber-attacks pose a new threat to V2X communications.Source: §5.3.2 · p. 21
5.3.3 Driving States Prediction/Training of Non-connected Vehicles.
- Predicting manual driving states within a reasonable horizon can improve trajectory optimization, while driver-assistance systems may train or direct people toward more predictable driving patterns.Non-connected and non-automated vehicles are treated as sources of uncertainty in control-input calculation.Source: §5.3.3 · p. 21
6 Conclusions
The conclusion reviews sensing and communication, human factors, and information-aware controller design as interrelated aspects of CAVs, with safety, operational efficiency, and user comfort in view.
6 Conclusions
- The review identifies heterogeneous sensing and communication technologies, their architectures and protocols, and future technologies as elements for CAV design.The authors emphasize fully utilizing heterogeneous sensing and communication to support safety, operational efficiency, and user comfort.Source: §6 · p. 21
- User-acceptable CAV design must consider comfort, preferences, reliability, behavioral adaptation, and situation awareness.Future work identified includes preference modeling, system-failure handling, and integration of multiple assistance systems.Source: §6 · p. 21
- The controller review covers route planning, driving-mode selection, and driving-mode execution with V2X; future work should consider centralized versus distributed control and mixed traffic.The mixed-traffic issues include manually driven and non-connected vehicles.Source: §6 · p. 21
- The conclusion states that CAVs can improve roadway safety and operational efficiency by reducing human errors in manual driving.Source: §6 · p. 21