Li Jinglu, Xiaohua Sun (2023). Exploring Proactivity in Human-Vehicle Interaction: Insights for proactive interaction Design. In: Human Factors in Transportation. AHFE Open Access, vol. 95, pp. 182–193. https://doi.org/10.54941/ahfe1003804
Artificial Intelligence (AI) is rapidly evolving, and systems are becoming more proficient in collaborating with humans. These systems not only respond to users' requests but can also anticipate their potential needs and meet them. However, the understanding of proactivity in the context of human-vehicle interaction is limited, and existing research is based on common sense rather than a clear concept of proactivity. In this study, we explore the theme and concept of proactivity in human-vehicle interaction through a literature review and case study. We also provide insights about the content that needs to be designed for proactive interaction. The study's findings will help researchers and designers better comprehend proactive interaction in the new relationships between humans and vehicles and support interaction design in this field.
The aim of this paper is to explore the concept of proactive interaction from existing literature and illustrate its application in proactive interaction design. By doing so, we aim to enhance the comprehension of proactive interaction design. Furthermore, we intend to identify the content of human vehicle pro- active interaction, which can be leveraged to create effective and efficient proactive interaction designs.
THE RELATED WORKS OF PROACTIVITY To explore the means of proactivity in human-vehicle interaction, we fol- low the prior study in many different disciplines, including human-computer interaction(HCI), human-robot interaction(HRI), human-vehicle interaction, and occupational psychology. It’s worth understanding one term from diffe- rent views in different contexts. We choose 20 critical papers from hundred related papers for inductive analysis.
We summarize information from a set of research papers, and categories them including author, year, context, description, discipline, and keywords in Table 1. The context is the proactivity-related context in the paper. The description is the proactivity-related description in the paper. Keywords are
understanding and predicting of behaviors, we need
to declare proactive construct.
Occupational psychology
Table 1. Continued.
Table available in the original PDF paper.
No.Authorsyear Context Description Key words
17(Bateman and Crant, 1993) This study investigated a personal disposition
18(Crant, 2000) This work describe four constructs related to
19(Parker and Collins, 2010) This work develop a framework designed to
20(Parker and Collins, 2010) This work clarify the similarities, differences, and
Exploring Proactivity in Human-Vehicle Interaction 187
refined from the proactivity-related descriptions, it’s also an original label of the proactivity concept inductive analysis.
DEVELOPING CONCEPT OF PROACTIVE INTERACTION The Theme of Proactivity
We follow the coding process in inductive analysis (Thomas, 2003). The spe- cific segments are the proactivity-related description from the papers. Then, label the segments to create categories, which is the keywords in Table 2. Reduce overlap and redundancy among those categories to 18 categories, including context, human supervised, autonomy action, explanatory, take autonomous action, relevance, user agent, different from react, function list, understandable, anticipation, initiate by system(robot), impact, user’s needs, user preferences, action in advance, anticipatory action, metrics. Last, create a model incorporating most important categories to create five sum- mary categories, which is consider and manage context, anticipation and predictive ability, autonomy and action, understandability and transparency, proactive functions. 01 Consider and manage context: the system takes into account the user’s context, including their environment, device, activity, and task, to perception and comprehension of context, to support and inform its actions. 02 Autonomy and action: the system’s ability to use predictive analytics to anticipate user needs and preferences and initiate actions to impact others. 03 Anticipation and predictive ability: the system is designed to anticipate the user’s needs and preferences and take anticipatory actions, using predictive analytics and other technologies to improve its accuracy. 04 Understandability and transparency: in collaboration between the user and the system, the system should make its actions transparent and under- standable to the user, including providing feedback and explanations in plain language. In order to build trust and promote efficient collaboration, the system must provide transparent feedback and explanations that help the user understand what actions are being taken and why.
The 02 and 03 both concepts involve the system’s ability to use predictive analytics to anticipate user needs and preferences and take proactive actions. Autonomy and action emphasize the initiation action part; Anticipation and
Table 2. First categories and summary categories for inductive analysis.
Table available in the original PDF paper.
First categories Summary categories
Anticipation and predictive ability
Context, relevance, impact Consider and manage context Different from react, action in advance, anticipation, anticipatory action Autonomy and action Human supervised, take autonomous action, autonomy action, user agent, initiate by system(robot)
Understandability and transparency
Explanatory, understandable, metrics, user’s needs, user preferences Function list Proactive functions
188 Li and Sun
predictive ability emphasis understanding the reasons behind the initiation of the action and the contents and ways of the specific action that should be taken to meet the user’s needs and preferences.
The functions of proactivity are an important part of proactive interaction research, not belong to the theme of the proactivity concept.
Concept of Proactive Interaction in Autonomous Vehicles Following the theme of proactivity, we propose the concept of proactivity, which refers to the ability of a system to take initiative and act without prior prompting, based on an understanding of the user’s goals and the context in which they are operating. This can involve the system offering advice or taking control in a situation where it is deemed necessary to enhance safety, efficiency, or convenience.
To achieve this, the system must anticipate the user’s needs and prefere- nces based on their previous behavior and the current situation. This allows the system to provide personalized recommendations or assistance that are tailored to the user’s goals and preferences.
Furthermore, the system should communicate its knowledge of the user’s goals and the environmental context in a transparent and understanda- ble way, which can help to build trust and enhance the user’s engagement with the system. This is particularly important in the case of autonomous vehicles, where safety and trust are critical factors in user acceptance and adoption.
THE CASES OF PROACTIVE INTERACTION DESIGN IN VEHICLE MARKET Many brands now offer a range of features designed to improve safety and convenience. Tesla’s Autopilot system, Ford, Volkswagen Front Assist system, General Motors’ Super Cruise system, and Toyota’s Safety Sense system offers a range of proactive features, such as automatic braking to avoid collisions, pedestrian detection, proactive lane-keeping, automatic lane changing and speed control, automatic cruise control, smart parking that find available parking spaces and guide the driver to them. These features can improve safety and convenience by alerting drivers to potential hazards and taking action to prevent accidents.
Although proactive features have been shown to offer several benefits, it is important to conduct further research to determine how these features impact user experiences. Given the highly complex nature of the driving context, particularly with higher levels of autonomy, there may be greater scope for the design of proactive features that extend service-related content, such as actions based on users’ preferences and state, rerouting based on users’ goals and events, and other similar functions.
In some of its models, the BMW Vision iNEXT is equipped with a digital assistant that can help passengers with a variety of tasks, such as boo- king restaurant reservations, ordering groceries, and finding parking spots. MBUX (Mercedes-Benz User Experience) system learns user behavior based
Exploring Proactivity in Human-Vehicle Interaction 189
on sensor data and event and location information. The car will recom- mend these activities to the user at the appropriate time. HANA (Honda Automated Network Assistant) determines the driver’s current emotional state by analyzing body language, physical readings, and current actions. HANA can measure the driver’s emotions and provide music recommendati- ons or suggestions, as well as other media and entertainment options based on their mood. “Xiaodu Car OS Intelligent Scene Engine” developed by Baidu can recognize the current scene and suggest services based on that scene, such as recommending playlists, providing automatic appointment reminders for upcoming maintenance times, and implementing facial recognition automatic payment in payment scenarios.
Those many cases in the market take the form of a wide range of functions and services, the emphasis on the initiation of action is predictions of user’s needs in context is an important shifting and orientation.
THE CONTENTS OF PROACTIVE INTERACTION DESIGN IN AUTONOMOUS VEHICLE According to the concept and the functions and services related features, the design contents can be constructed. There are 3 parts involved. Anticipation of goal and need based on scenario, proactive behavior based on proactivity functions, and context-aware communication based on collaboration
Anticipation of Goal and Need Based on Scenario The driving scenario is the space where human-vehicle interaction occurs. Bringing the proactive interaction behavior of AVs into specific scenarios for analysis is a precondition for design and research. A typical scenario is one that is representative and has a high probability of occurrence in driving context.
The study recruited volunteers who were licensed drivers with at least two years driving and who drove frequently, with more than 50% of their weekly trips being by car. The volunteers were asked to use a GoPro camera to film a 15–30 minute driving video based on their daily activities. The driving videos we collected were primarily filmed in Shanghai, China, and San Francisco, California, the US. totaling 2 hours and 30 minutes in duration.
The collected video materials were then analyzed to classify the driving scenarios into purpose-driven scenarios, which have clear goals and involve activities such as commuting, traveling, and running errands, and function- driven events, which are fragmented and include tasks such as finding parking spaces, refueling, or answering phone calls. By analyzing the scenarios, the researchers identified the users’ goals and need in various driving scenarios, which helped establish a design target in proactive interaction design. The user goals of each scenario can be summarized in Table 3. These goals encom- pass eight types: information acquisition, problem-solving, making plans, executing plans, routine activities, status monitoring, security assistance, and information entertainment. Designers can refer to these effectively locate the user’s needs and align user goals at design time.
190 Li and Sun
Table 3. The goals of (1) purpose-driven scenarios and (2) function-driven events. (1) Purpose-driven scenarios User goals
Table available in the original PDF paper.
Commuting Get to work on time and relax on the way home. Business trips Arrive at the destination and avoid emergencies. Going to grocery Arrive at the destination and find an available parking lot. Traveling Arrive at the destination and enjoy the scenery along the way. Going to the hospital Arrive before the appointment time and find a parking space. Visiting friends Arrive at a friend’s house at the appointed time.
(2) Function-driven events User goals Traffic jam Relieve anxiety, get traffic information factually Play music Quickly find the target music and choose music according to the context Find a parking space Quickly find an available parking space that suits the situation (free/ close to an exit/…) Find a restaurant Find a favorite one/ the most time-saving one/… Get gas Plan your trip reasonably and find the right gas station Pedestrians across Drive safely, avoid danger Schedule change Update the schedule, don’t delay the original plan Route change Plan a better route without delaying the itinerary Cart approaching Drive safely and keep a safe distance Vehicle wear/maintenance information Keep your vehicle in good condition before something goes wrong Extreme weather Drive safely, avoid danger
Proactive Behavior Based on Proactivity Functions The autonomy of proactive interaction refers to the ability of a system to independently take actions. The degree of autonomy in proactive interaction can be manifested in the strength of the system’s ability to act independently during the execution of tasks. This can be implemented in design through the integration of proactive functions.
In human-vehicle cooperation, the proactive autonomy functions of inte- raction can be classified into seven types based on the different behaviors exhibited by intelligent vehicles. These types include push, remind, recom- mend, suggest, inquire, request, and execute (report or silent). Each type comprises several application contents, as outlined in Table 4. These contents can be developed into design cases to improve design efficiency.
Context-Aware Communication Based on Collaboration The degree of situational awareness conveyed to users is a critical dimen- sion of communication between users and systems, serving as an indicator of the system’s ability and as a basis for establishing trust. When the system supports or violates their goals and attracts users’ attention, it will positively or negatively affect their acceptance, user experience, and trust. In contrast to literature that focuses on macro-technical planning(Omeiza
Exploring Proactivity in Human-Vehicle Interaction 191
Table 4. Autonomy function type and application contents in AVs.
Table available in the original PDF paper.
Autonomy function type Application contents Information push News information, service information, managed information Recommended Points of interest(POI), infotainment content, system applications Reminder Danger prediction, event prediction, information service Ask Abnormal state ask, event ask, plan ask Requests Functional requests, permission requests Suggestions Locations suggest, user behavior suggest, plan suggest, security operations suggest Execution (reporting/silence) Emergency decision-making, safety decision-making, safety supervision, service on the road decision-making, in-vehicle environment adjustment, functional autonomous execution
et al., 2021)(Atakishiyev et al., 2021), this study pays more attention to the relationship between system behavior and user experience.
The proactive interactions in context-aware communication involve infor- mative and explanatory information. Informative information presents the vehicle system’s situational awareness results and proactive behaviors, while explanatory information provides content-based, reason-based, and intera- ctive presentations regarding the proactive behavior and its contents. Infor- mative information comprises changes in user status, environmental status, and task status.
CONCLUSION
The concept of human-vehicle proactive interaction has gained significant attention in recent years due to its potential to improve road safety and enh- ance the driving experience. This approach enables vehicles to anticipate the driver’s intentions and respond in advance, using advanced sensors, algo- rithms, and communication systems. In this work, we sum up the theme of proactive interaction by inductive analysis of literature and define the concept of proactive interaction in AVs. The function-based and service-based cases are the directions we can follow. As a result, the content of proactive inte- raction design in AVs is the early guideline to support proactive interaction design in the process.
REFERENCES
Atakishiyev, S. et al. (2021) ‘Explainable Artificial Intelligence for Autonomous Driving: A Comprehensive Overview and Field Guide for Future Research Dire- ctions’, arXiv preprint arXiv:2112.11561 [Preprint]. Bader, R. (no date) ‘Proactive Recommender Systems in Automotive Scenarios’,
p. 348. Bader, R., Woerndl, W. and Prinz, V. (2010) ‘Situation Awareness for Proactive In-Car Recommendations of Points-Of-Interest (POI)’, p. 10. Bateman, T. S. and Crant, J. M. (1993) ‘The proactive component of organizational
behavior: A measure and correlates’, Journal of organizational behavior, 14(2), pp. 103–118.
192 Li and Sun
Cesta, A. et al. (2007) ‘Proactive Assistive Technology: An Empirical Study’, in C. Baranauskas et al. (eds) Human-Computer Interaction – INTERACT 2007. Berlin, Heidelberg: Springer (Lecture Notes in Computer Science), pp. 255–268. Available at: https://doi.org/10.1007/978-3-540-74796-3_25. Crant, J. M. (2000) ‘Proactive behavior in organizations’, Journal of Manage-
ment, 26(3), pp. 435–462. Available at: https://doi.org/10.1016/S0149-2063(00) 00044-1. Isbell, C. L. and Pierce, J. S. (2005) ‘An IP Continuum for Adaptive Interface Design’, p. 10. Kabtoul, M., Spalanzani, A. and Martinet, P. (2020) ‘Towards Proactive Navigation:
A Pedestrian-Vehicle Cooperation Based Behavioral Model’, in ICRA 2020 - IEEE International Conference on Robotics and Automation. Paris, France (IEEE Inter- national Conference on Robotics and Automation Proceedings), pp. 6958–6964. Available at: https://hal.archives-ouvertes.fr/hal-02509637 (Accessed: 4 May 2022). Kraus, M. et al. (2020) ‘A Comparison of Explicit and Implicit Proactive Dialogue Strategies for Conversational Recommendation’, in Proceedings of The 12th Language Resources and Evaluation Conference. LREC 2020, Marseille, France: European Language Resources Association, pp. 429–435. Available at: https://www.aclweb.org/anthology/2020.lrec-1.54 (Accessed: 20 August 2020). Mok, B. K.-J. et al. (2015) ‘A place for every tool and every tool in its place: Perfor-
ming collaborative tasks with interactive robotic drawers’, in 2015 24th IEEE International Symposium on Robot and Human Interactive Communication (RO-MAN). IEEE, pp. 700–706. Nothdurft, F., Ultes, S. and Minker, W. (no date) ‘Finding Appropriate Interaction Strategies for Proactive Dialogue Systems—An Open Quest’, p. 8. Omeiza, D. et al. (2021) ‘Explanations in autonomous driving: A survey’, IEEE
Transactions on Intelligent Transportation Systems [Preprint]. Parker, S. K. and Collins, C. G. (2010) ‘Taking stock: Integrating and differentiating multiple proactive behaviors’, Journal of management, 36(3), pp. 633–662. Peng, Z. et al. (2019) ‘Design and Evaluation of Service Robot’s Proactivity in
Decision-Making Support Process’, in Proceedings of the 2019 CHI Conference on Human Factors in Computing Systems - CHI’19. the 2019 CHI Conference, Glasgow, Scotland Uk: ACM Press, pp. 1–13. Available at: https://doi.org/10. 1145/3290605.3300328. Salovaara, A. and Oulasvirta, A. (2004) ‘Six Modes of Proactive Resource Management: A User-Centric Typology for Proactive Behaviors’, p. 4. Sirithunge, C., Jayasekara, A. G. B. P. and Chandima, D. P. (2019) ‘Proactive Robots
With the Perception of Nonverbal Human Behavior: A Review’, IEEE Access, 7, pp. 77308–77327. Available at: https://doi.org/10.1109/ACCESS.2019.2921986. Smirnov, A., Shilov, N. and Gusikhin, O. (2015) ‘Socio-cyberphysical system for proactive driver support approach and case study’, in 2015 12th International Conference on Informatics in Control, Automation and Robotics (ICINCO). 2015 12th International Conference on Informatics in Control, Automation and Robotics (ICINCO), pp. 289–295. Tan, H. et al. (2020) ‘Relationship between social robot proactive behavior and the
human perception of anthropomorphic attributes’, Advanced Robotics, 34(20), pp. 1324–1336. Available at: https://doi.org/10.1080/01691864.2020.1831699. Tennenhouse, D. (2000) ‘Proactive computing’, Communications of the ACM, 43(5), pp. 43–50. Available at: https://doi.org/10.1145/332833.332837. Thomas, D. R. (2003) ‘A general inductive approach for qualitative data analysis’.
Exploring Proactivity in Human-Vehicle Interaction 193
Woerndl, W. et al. (2011) ‘A model for proactivity in mobile, context-aware recommender systems’, in Proceedings of the fifth ACM conference on Recommender systems. New York, NY, USA: Association for Computing Machinery (RecSys’ 11), pp. 273–276. Available at: https://doi.org/10.1145/2043932.2043981. Xiao, J., Catrambone, R. and Stasko, J. (2003) ‘Be Quiet? Evaluating Proactive and
Reactive User Interface Assistants’, p. 10.
LISTED IN THE AHFE DIGITAL BOOKSTORE CATALOGUE
Explore the complete digital volume
Human Factors in Transportation
Read this individual research paper free through the original publisher, or explore the full digital proceedings book.
ISBN 978-1-958651-71-1
Explore this ISBN in the AHFE Bookstore ↗Listed in the July 2026 bookstore catalogue; search by ISBN to confirm current availability · Digital book purchase optional