How can I improve my performance here?Figure 1: After completing a lap, the driver pauses the video at a specific segment and asks ApexTrainer for feedback.The system generates context-aware suggestions based on the referenced location.

Motor skills education often requires experienced professionals who can provide personalized instruction. Unfortu-nately, the availability of high-quality training can be limited for specialized tasks, such as high performance racing. Several recent works have proposed AI -assistance for motor skills instruction, ranging from rehabilitation to surgical robot tele-operation. However, these works often make simplifying assumptions on the student learning process, and fail to model how a teacher's assistance interacts with different individuals' abilities when determining optimal teaching strategies. Inspired by the idea of scaffolding from educational psychology, we leverage shared autonomy, a framework for combining user inputs with robot autonomy, to aid with curriculum design. Our key insight is that the way a student's behavior improves in the presence of assistance from an autonomous agent can highlight which sub-skills might be most “learnable” for the student, or within their Zone of Proximal Development. We use this to design Z-COACH, a method for using shared autonomy to provide personalized instruction targeting interpretable task sub-skills. In a user study$(\mathrm{n}=50)$, where we teach high performance racing in a simulated environment of the Thunderhill Raceway Park with the CARLA Autonomous Driving simulator, we show that Z-COACH helps identify which skills each student should first practice, leading to an overall improvement in driving time, behavior, and smoothness. Our work shows that increasingly available semi-autonomous capabilities (e.g. in vehicles, robots) can not only assist human users, but also help teach them.

Learning motor skills for sports or performance driving is often done with professional instruction from expert human teachers, whose availability is limited. Our goal is to enable automated teaching via a learned model that interacts with the student similar to a human teacher. However, training such automated teaching systems is limited by the availability of highquality annotated datasets of expert teacher and student interactions as they are difficult to collect at scale. To address this data scarcity problem, we propose an approach for training a coaching system for complex motor tasks such as high performance driving via a Multi-Task Imitation Learning (MTIL) paradigm. MTIL allows our model to learn robust representations by utilizing self-supervised training signals from more readily available non-interactive datasets of humans performing the task of interest. We validate our approach with (1) a semi-synthetic dataset created from real human driving trajectories, (2) a professional track driving instruction dataset, (3) a track-racing driving simulator human-subject study, and (4) a system demonstration on an instrumented car at a race track. Our experiments show that the right set of auxiliary machine learning tasks improves prediction of teaching instructions. Moreover, in the human subjects study, students exposed to the instructions from our teaching system improve their ability to stay within track limits, and show favorable perception of the model's interaction with them, in terms of usefulness and satisfaction.

People's goal-directed behaviors are influenced by their cognitive biases, and autonomous systems that interact with people should be aware of this. For example, people's attention to objects in their environment will be biased in a way that systematically affects how they perform everyday tasks such as driving to work. Here, building on recent work in computational cognitive science, we formally articulate the \textit{attention-aware inverse planning problem}, in which the goal is to estimate a person's attentional biases from their actions. We demonstrate how attention-aware inverse planning systematically differs from standard inverse reinforcement learning and how cognitive biases can be inferred from behavior. Finally, we present an approach to attention-aware inverse planning that combines deep reinforcement learning with computational cognitive modeling. We use this approach to infer the attentional strategies of RL agents in real-life driving scenarios selected from the Waymo Open Dataset, demonstrating the scalability of estimating cognitive biases with attention-aware inverse planning.

Personalization of autonomous vehicles (AVs) may significantly increase acceptance. In particular, we hypothesize that the similarity of an AV's driving style compared to a user's driving style, the level of aggressiveness of the driving style, and other subjective factors (e.g., personality) will have a major impact on user's willingness to use the AV. In this work, we 1) develop a data-driven approach to personalize driving style and calibrate the level of aggressiveness and 2) investigate the subjective factors that impact user preference. Across two human subject studies (n = 54), we demonstrate that our approach can mimic the driving styles and tune the level of aggressiveness. Second, we leverage our framework to investigate the factors that impact homophily. We demonstrate that our approach generates driving styles objectively ($p < .001$) and subjectively ($p = .002$) consistent with end-user styles ($p < .001$) and can effectively isolate and modulate a dimension of style (i.e., aggressiveness) ($p < .001$). Furthermore, we find that personality ($p < .001$), perceived similarity ($p < .001$), and high-velocity driving style ($p = .0031$) significantly modulate the effect of homophily.

Should fully autonomous vehicles (FAVs) be designed inclusively and accessibly, independence will be transformed for millions of people experiencing transportation-limiting disabilities worldwide. Although FAVs hold promise to improve efficient transportation without intervention, a truly accessible experience must enable user input, for all people, in many driving scenarios (e.g., to alter a route or pull over during an emergency). Therefore, this paper explores desires for control in FAVs among (n=23) people who are blind and visually impaired. Results indicate strong support for control across a battery of driving tasks, as well as the need for multimodal information. These findings inspired the design and evaluation of a novel multisensory interface leveraging mid-air gestures, audio, and haptics. All participants successfully navigated driving scenarios using our gestural-audio interface, reporting high ease-of-use. Contributions include the first inclusively designed gesture set for FAV control and insight regarding supplemental haptic and audio cues.

This work presents a novel ultrasonic haptic interface to improve nonvisual perception and situational awareness in applications such as fully autonomous vehicles. User study results (n=14) suggest comparable performance with the dynamic ultrasonic stimuli versus a control using static embossed stimuli. The utility of the ultrasonic interface is demonstrated with a prototype autonomous small-scale robot vehicle using intersection abstractions. These efforts support the application of ultrasonic haptics for improving nonvisual information access in autonomous transportation with strong implications for people who are blind and visually impaired, accessibility, and human-in-the-loop decision making.

The Children's Risk Utility Measure (CRUM) is an interactive, modified version of the Balloon Analogue Risk Task (BART) that is appropriate for young children. This paper utilizes a Bayesian IRT model (an extension of logistic regression) to examine the impact of subject-level characteristics (age, working memory, inhibitory control) and trial-level characteristics (e.g. block, order, etc.) on the performance of children ages 3-6 years on the CRUM. The Bayesian model is especially helpful when performing IRT models-which have many parameters-on small data sets. The order of the trials was not associated with lower difficulty, indicating that children's performance remained steady as the task went on. Working memory performance was negatively associated with children's latent CRUM performance, while high scores on an inhibitory control task were positively related to child's latent CRUM performance. Age and the interaction of age with working memory performance were both negligible effects. The IRT model was an effective way to estimate latent CRUM performance.