Enhance user experience as more car-associated systems emerge, self-driving cars are becoming a reality. Studies show that not having control can negatively affect the user experience (UX) in busy urban areas. To improve this, giving suitable feedback is necessary for a better experience. Including various feedback methods is crucial in designing the overall user experience for AVs. Through feedback modalities like light, audio, text, visualization, and vibration, passengers can better understand the actions of self-driving cars. We explore autonomous driving absorption. They emphasize the importance of clear connection through different feedback types for a better overall user experience by being aware of and communicating the vehicle’s intentions. We aim to improve the idea and pleasure of passengers in autonomous cars.
Enhance user experience within autonomous vehicles
Related work:
enhance user experience to Explaining complex artificial systems to people is the goal of reasonable artificial intelligence research. This field, painted in various studies, centers on making AI readable and focuses on providing explanations. A challenge arises when AI experts do not need to improve at clarifying these intricate models. To address this, there is a need for simpler, context-dependent. And well-timed ways of disclosing with humans. This explores the connection between User Experience Design (UXD) and XAI to enhance clearness in human-AI interaction. In the context of semi-autonomous driving, recent research has looked into using feedback methods. Like light, audio, meditations, text, or vibration to convey the AI’s awareness and intent to drivers. At the same time, these methods have proven helpful in increasing view and trust. No study has focused on the user experience in autonomous driving concerning AI communication. This article takes a step in that direction.
Prototype:
Driving situation:
There are various self-driving situations to enhance user experience. We identified four main categories in a workshop involving researchers and automotive experts. Proactive and reactive situations, each falling into either critical or noncritical scenarios. Fierce situations involve the vehicle getting ahead and preparing for potential challenges. While reactive, ones involve responding to unexpected events. Proactive and reactive situations can be severe, demanding fast attention, or noncritical, less urgent. This category helps in understanding and addressing different types of challenges. That autonomous vehicle may encounter on the road.
Proactive noncritical:
In which the AV has ample time to react without putting human health and lives at risk. For instance, if there’s a small creation site obstructing the vehicle’s path. This is the way enhance user experience the AV can safely and soundly wait for oncoming traffic to pass before crossing around the construction site. This setting highlights the AV’s ability to handle less urgent challenges easily. It is ensuring the welfare of both passengers and others on the road. This proactive approach allows the AV to navigate obstacles efficiently. It is contributing to a smooth and secure autonomous driving experience.
Proactive critical:
In critical situations where human health or lives are at risk. The autonomous vehicle is designed to react sharply. For instance, if a car in on-coming traffic is overtaking another vehicle. The AV can quickly recognize the danger and apply brakes in advance, preventing a possible unsafe condition. This proactive response showcases the AV’s ability to prioritize safety. Proving its capacity to make sensible decisions to avert risks. By acting rapidly in critical moments, the AV ensures the well-being of passengers and others on the road. It emphasizes its promise to enhance safety in autonomous driving scenarios.
Reactive noncritical:
In reactive, noncritical situations, AV faces challenges with limited reaction time. But the problem is not immediately life-threatening. For instance, if a dog suddenly appears behind a parked car and dashes across the street, the Must reacts quickly. In this setup, it performs an emergency brake to avoid any harm. While the situation does not expose human health, the AV promptly addresses unexpected events. It showcases its ability to handle noncritical yet rapidly unfolding situations and ensures passengers’ safety and potential road problems.
Reactive critical:
Enhance user experience In reactive critical positions, the autonomous vehicle faces pending danger with limited reaction time. For example, if a car in oncoming traffic overtakes another vehicle, the AV must act swiftly to prevent a possible accident. In this scenario, the AV performs an emergency brake, showcasing its ability to respond urgently to life-threatening conditions. While the AV may not have enough time for proactive measures, its quick reaction helps avert a collision. It is prioritizing the safety of both passengers and others on the road. This highlights the importance of the AV’s capability to address critical situations promptly.
Feedback design:
A virtual reality VR prototype was made using Unity. Various feedback methods were designed: light, visualizations, text, audio, and vibration. Each provides different levels of information. These elements aim to enhance user experience by offering diverse and meaningful feedback in the VR environment.
Light:
The light feedback in the VR prototype employs a traffic light system inspired by prior research. Two light bars are located on the vehicle’s front left and right. Serve as indicators of the driving status and reactions. When everything is normal, the lights are white. If a problem is recognized without immediate response, the lights turn green. Yellow indicates recognition of an obstacle with planned braking. While red signals an identified risk of collision, prompting emergency braking. If an event occurs on the left side, the left light bar changes color and the same happens on the right side. Both light bars change color if an event unfolds directly before the vehicle. They are enhancing the communication of driving situations to the VR user.
Audio:
The audio feedback in the VR prototype uses abstract sounds, known as auditory icons or ear cons, to talk about different situations to passengers. For the “Continue drive” setup, a brief and soothing two-note sound signal that the shuttle is continuing its drive creates a relaxed atmosphere. In the “Reaction” situation, a neutral two-note ring indicates that the vehicle has recognized a scenario and is preparing to react and keeping passengers informed without causing alarm. For potential dangers, the “Warning” situation is communicated through three consecutive alerting beep sounds, creating a sense of urgency. These audio cues aim to provide passengers with clear and distinct signals for various driving situations. They contribute to their understanding and overall experience in the virtual reality environment.
Visualization:
The visualization aspect of the VR prototype draws inspiration from miniature Tesla and Waymo likenesses. Displayed to passengers, it showcases familiar objects, vehicles, walkers, and route information. This visual aid enhances passenger awareness, providing a clear and informative portrayal of the virtual environment. Real-world notion techniques from well-known automotive companies influence it.
Text:
The Virtual reality sample discusses necessary driving actions through short text messages. The emphasis is on why messages follow the approach introduced, where testers showed improved driving performance. For the four driving situations in the model, the text messages daily are displayed as a construction site, overtaking, animal, and ghost driver. The uniform presentation of these messages aims to provide a clear and constant visual indication. They allow passengers to quickly grasp the driving situations, foster better understanding, and align with the proven potency of the “why-message” concept in enhancing user ability.
Vibration:
In the Virtual Reality model, seat vibrations play a role in communicating events. The seat vibrates three times for proactive situations, each lasting 0.3 seconds. This pattern signals preparation for upcoming scenarios. In reactive situations, the seat vibrates six times, with a shorter duration of 0.1 seconds per vibration. This more intense and rapid vibration indicates an instant response to unexpected events. By using distinct vibration patterns, the system provides passengers with a concrete sense of the driving context. They are enhancing their awareness and understanding of proactive and reactive situations in the virtual environment.
Driving simulator:
A virtual scene was set in an urban area with a single lane for each driving position. The AV moved at about 30 km/h, and experiencing each situation took roughly 25 seconds. Except for vibration, all feedback types were included in the virtual world. The vibration feedback used a motor connected to an Arduino Uno, conversing with the Unity prototype. This motor was placed inside a small metal tube on the seat. This setup allowed passengers to feel vibrations, adding a physical layer to the virtual experience and making it more immersive and informative.
Study design:
Hypothesis:
The hypothesis proposes using different feedback types. Light, audio, visualization, text, and vibration will positively impact passenger UX in various driving situations. The experiment used a mixed-method within-subjects design. All 22 participants experienced four driving effects of each feedback modality. The order of practices is balanced using a Latin square design. Participants, unknown of the modalities, rated the pragmatic and hedonic quality of UX with the UEQ-S. Baseline rides without feedback served as a comparison. Testers gave feedback through a question, think-aloud method, and unstructured interviews. They asked to specify preferred feedback modalities for different situations. With diverse backgrounds and no autonomous driving experience, participants underwent a one-hour test. They were ensuring a complete evaluation of the impact of varied feedback on passenger UX.
Limitations:
To enhance user experience the study has limitations. While VR created emotional reactions, this is an image lacking real-world immersion. Unlike real scenarios where passengers multitask, testers were instructed to focus on the ride. All testers were first-time users, limiting familiarity, and the study included a specific age group. The digital world’s limits, like graphics and sound, may affect the overall experience. Despite these limitations, the experiment provides valuable insights into the impact of different feedback modalities on passenger UX in autonomous driving scenarios.
