As robots enter our homes and workplaces, they will have more direct access to us, our information, and our daily lives. The more a robot knows about us, the more helpful it can be, at least in theory. Where is the balance between privacy and utility? How might the robot communicate the tradeoffs between privacy and utility in a nuanced but clear manner? How can we ensure that the perception of privacy-protection offered by the robot accurately reflects how much the robot is actually preserving our privacy? How can a robot learn information that it can use to better help us without being intrusive? What behaviors should the robot have (or not have) to not only ensure that it protects our privacy, but also is perceived as protecting our privacy? How can we measure the effectiveness of these behaviors, so that we can track our progress towards more useful, less invasive robot companions? This workshop is the the third in a series at HRI, and will bring together researchers from a wide variety of intellectual communities to look at these questions, identify promising research directions, and set an agenda for how to start making progress. We are particularly interested in expanding the core of researchers interested in privacy-related issues in HRI, and in those "privacy-curious'' researchers who want to find out more about he intersection of privacy and their own research.

Privacy is an important yet understudied focus in consideration of successful human-robot interaction (HRI). In this paper, we present a thematic analysis of the topics discussed in the Privacy-Aware Robotics Workshop held at the HRI Conference in 2024. The analysis points across the perspectives of User, Engineering, and Society with particular identification of open "interdisciplinary zones" at the intersections of these perspectives. Based on that, we formulate and present three main themes for future research directions: the tension between robot functioning and effectiveness of privacy implementations in real-world contexts; the need to involve and empower target user communities in the co-design of privacy-aware robots, and the consideration of regulatory frameworks that extend across jurisdictions in robot design.

Ultraviolet-C (UV-C) robot irradiation is a promising approach for disinfecting surfaces contaminated by pathogens in healthcare settings. However, limitations exist with current UV disinfection robots, including coverage for complex surface geometries. This research presents a system for human-guided robotic UV disinfection that uses empirical sensor measurements rather than relying on high-accurate models for UV map coverage. Human guidance is integrated into the methodology to enhance disinfection, aiding in addressing complex shaped objects and topologies. Further, a validation test confirmed that our estimation approach reliably underestimates the UV exposure, which is beneficial for ensuring thorough disinfection. Initial pilot studies demonstrated that while autonomous disinfection was effective for simple objects like tabletops, human-guided disinfection, especially with feedback, improved coverage and speed for complex shapes like mugs. Combining human intuition with autonomy shows promise for enhancing robotic disinfection effectiveness.

Ultraviolet-C (UV-C) robot irradiation is a promising approach for disinfecting surfaces contaminated by pathogens in healthcare settings. However, limitations exist with current UV disinfection robots, including coverage for complex surface geometries. This research presents a system for human-guided robotic UV disinfection that uses empirical sensor measurements rather than relying on high-accurate models for UV map coverage. Human guidance is integrated into the methodology to enhance disinfection, aiding in addressing complex shaped objects and topologies. Further, a validation test confirmed that our estimation approach reliably underestimates the UV exposure, which is beneficial for ensuring thorough disinfection. Initial user studies demonstrated that while autonomous disinfection was effective for simple objects like tabletops, human-guided disinfection, especially with feedback, improved coverage and speed for complex shapes like mugs. Combining human intuition with autonomy shows potential for enhancing robotic disinfection effectiveness.

As robots become increasingly active around human environments, they must navigate both physical and social realms, necessitating awareness of their surroundings and inhabitants, including the potential collection of sensitive data. To be accepted in human spaces, they need to be trusted to handle personal information adequately, not only adhering to security and data protection standards, but also aligning with contextual norms, individual expectations, and domain-specific requirements. Challenges intensify when robots engage with multiple humans across varied contexts over extended periods. Drawing from psychology, sociology, ethics, and law, and the experience of participants, we seek to outline dimensions and prerequisites for privacy-awareness in HRI. In this workshop we want to discuss methodologies, user interfaces, and personalizing options, and AI reasoning to design privacy-aware robot behavior in the human-robot interaction community.

Home robots operate in diverse and dynamic environments, delivering a range of functions that enhance utility. Many of these functions span extended periods, from weeks to months, typically improving through observations and interactions. Efficient development and validation of these functions necessitate simulations that can run faster than real time. However, many current robot simulators focus on high-fidelity physics simulation that limits their speed to a small multiple of real time. While these are tailored for critical low-level functions, they aren't optimized for simulating higher-level functions such as learning human behaviors and interacting with them. In this work, we introduce Fleet2D, a fast, lightweight simulator designed for long-term human-robot interactions in the home. By abstracting away low-level physics, aggressively caching compute-intensive operations, and operating in a simplified twodimensional world, we are able to perform realistic simulations of robot behavior at more than 10,000 times real time. We present the design, development, and validation of Fleet2D, showcase its effectiveness on a number of use cases in home robotics, and discuss how it has accelerated the development cycle for a home robot. Finally, we make Fleet2D open source for the research community at github.com/amazon-science/fleet2d.

For social robots to successfully integrate into daily life in home environments, they will need reliable models of the way people perceive and use space in the home. This paper explores the problem of obtaining annotated training data at scale for subjective judgments about spatial locations. Focusing on the use case of identifying good and bad parking spots for a social robot operating in a home environment, two experiments are presented. The first study shows that the presentation of context-rich 3D images to human annotators yields notably different outcomes from those obtained when using 2D robot navigation maps. We attribute the source of these differences to a set of features visible only in the 3D views and introduce a technique for labeling these features on the 2D maps. The second study reveals that using labeled 2D maps produces annotation data very similar to that obtained using 3D images. Since a labeled 2D map can be generated at a fraction of the cost of a full set of 3D views, we recommend this method as a scalable approach to collecting subjective spatial data annotations in everyday environments.

Many critical robot environments, such as healthcare and security, require robots to account for contextdependent criteria when performing their functions (e.g., navigation). Such domains require decisions that balance multiple factors, making it difficult for robots to make contextually appropriate decisions. Multi-Objective Optimization (MOO) methods offer a potential solution by trading off between objectives; however concepts like Pareto fronts are not only expensive to compute but struggle with differentiating among solutions on the Pareto front. This work introduces the Contextual Multi-Objective Path Planning (CMOPP) algorithm, which enables the robot to trade off different complex costs dependent on context. The key insight of this work is to separate the path planning and path cost estimation into two independent steps, thus significantly reducing computation cost without impacting the quality of the resulting path. As a result, CMOPP is able to accurately model path costs, which provide meaningful trade-offs when choosing a path that best fits the context. We show the benefits of CMOPP on case studies that demonstrate its contextual path planning capabilities. CMOPP finds contextually appropriate paths by first reducing the search space up to 99.9% to a near-optimal set of paths. This reduction enables the generation of accurate path cost models, using up to 90% less computation than similar methods.

Robots are being increasingly used in the fight against highly-infectious diseases such as Ebola, MERS, and SARS-CoV-2. Many of these robots use ultraviolet lights mounted on a mobile base to inactivate the pathogens. While the lights are generally effective at irradiating open spaces and walls, they are less effective when it comes to horizontal surfaces, because of the orientation of the light sources. This can be problematic for pathogens such as Ebola, where transmission via contaminated work surfaces, which are often horizontal, is a concern. In this paper, we describe the design, implementation, and testing of an ultraviolet light disinfection system implemented on a mobile manipulator robot designed to address the problem of horizontal surface disinfection. A human supervisor designates a surface for disinfection, the robot autonomously plans and executes an end-effector trajectory to disinfect the surface to the required certainty, and then displays the results for the supervisor to verify. We also provide some background information on Ultraviolet Germicidal Irradiation (UVGI) and describe how we constructed and validated models of ultra-violet radiation propagation and accumulation in our system. Finally, we describe our implementation on a Fetch mobile manipulation platform, and discuss how the practicalities of implementation on a real robot affect our models.

Recently, the robotics industry celebrated its 60-year anniversary. We have used robots for more than six decades to empower people to do things that are typically dirty, dull and/or dangerous. The industry has progressed significantly over the period from basic mechanical assist systems to fully autonomous cars, environmental monitoring and exploration of outer space. We have seen tremendous adoption of IT technology in our daily lives for a diverse set of support tasks. Through use of robots we are starting to see a new revolution, as we not only will have IT support from tablets, phones, computers but also systems that can physically interact with the world and assist with daily tasks, work, and leisure activities. The present document is a summary of the main societal opportunities identified, the associated challenges to deliver desired solutions and a presentation of efforts to be undertaken to ensure that US will continue to be a leader in robotics both in terms of research innovation, adoption of the latest technology, and adoption of appropriate policy frameworks that ensure that the technology is utilized in a responsible fashion.

Robots are being increasingly used to fight against the novel Coronavirus (SARS-CoV-2) in the current global pandemic. In previous work, we looked at the use of sophisticated mobile manipulation robots to perform surface disinfection in the context of the Ebola Virus Disease, using ultraviolet light. In this work, a human supervisor designates a surface to be disinfected. Then the robot autonomously plans a set of motions to disinfect the surface, using an ultraviolet light held in its gripper. Finally, the system displays the amount of ultraviolet radiation delivered to each part of the surface, so that the human supervisor can verify that the operation was successful. In this paper, we describe this work, and discuss how it could be adapted to the other pathogens, including SARS-CoV-2. In particular, we discuss how to move from disinfecting surfaces, which are of concern for Ebola, to open space disinfection, which is important in the fight against COVID-19.

Robots are being increasingly used in the fight against highly-infectious diseases such as the Novel Coronavirus (SARS-CoV-2). By using robots in place of human health care workers in disinfection tasks, we can reduce the exposure of these workers to the virus and, as a result, often dramatically reduce their risk of infection. Since healthcare workers are often disproportionately affected by large-scale infectious disease outbreaks, this risk reduction can profoundly affect our ability to fight these outbreaks. Many robots currently available for disinfection, however, are little more than mobile platforms for ultraviolet lights, do not allow fine-grained control over how the disinfection is performed, and do not allow verification that it was done as the human supervisor intended. In this paper, we present a semi-autonomous system, originally designed for the disinfection of surfaces in the context of Ebola Virus Disease (EVD) that allows a human supervisor to direct an autonomous robot to disinfect contaminated surfaces to a desired level, and to subsequently verify that this disinfection has taken place. We describe the overall system, the user interface, how our calibration and modeling allows for reliable disinfection, and offer directions for future work to address open space disinfection tasks.

Automated systems like self-driving cars and “smart” thermostats are a challenge for fault-based legal regimes like negligence because they have the potential to behave in unpredictable ways. How can people who build and deploy complex automated systems be said to be at fault when they could not have reasonably anticipated the behavior (and thus risk) of their tools? Part of the problem is that the legal system has yet to settle on the language for identifying culpable behavior in the design and deployment for automated systems. In this article we offer an education theory of fault for autonomous systems—a new way to think about fault for all the relevant stakeholders who create and deploy “smart” technologies. We argue that the most important failures that lead autonomous systems to cause unpredictable harm are due to the lack of communication, clarity, and education between the procurer, developer, and users of these technologies. In other words, while it is hard to exert meaningful control over automated systems to get them to act predictably, developers and procurers have great control over how much they test these tools and articulate their limits to all the other relevant parties. This makes testing and education one of the most legally relevant point of failures when automated systems harm people. By recognizing a responsibility to test and educate each other, foreseeable errors can be reduced, more accurate expectations can be set, and autonomous systems can be made more predictable and safer.

Advances in robotics technology will bring more teleoperated robots into homes to perform a variety of household tasks. This raises new privacy concerns as the remote operator can control the robot and its camera, and record its sensor data. One way to provide some privacy protection is through on-board processing of the data to filter out sensitive visual information. But what do people want hidden, and how should we hide it? Do the personality traits of a particular user influence that choice? We designed an 85-question survey to help answer these questions and analyzed the data from 81 respondents. We found that people are most concerned about hiding identifiable personal or financial information and valuables from a household robot, and we found that they prefer stronger filters to hide such items. We also found some evidence of the existence of correlations between a person’s familiarity with technology, sociability, and trust and their privacy concerns.

Background: Go Baby Go is a community program that provides modified ride-on cars to young children with disabilities. Aims: (1) To describe the real world modified ride-on car usage of young children with disabilities; (2) To compare subjectively reported modified ride-on car usage recorded by parents with objectively reported usage based on electronic tracking data. Methods: 14 young children (1-3 years old) with disabilities used a modified ride-on car for three months. Results: On average, parent-reported activity log data indicated that children used the modified ride-on car for 17.8 minutes per session (SD = 9.9) and 195.1 total minutes (SD = 234.8) over three months. Objective tracking data indicated 16.5 minutes per session (SD = 8.6) and 171.4 total minutes (SD = 206.1) over three months. No significant difference of modified ride-on car usage was found between parent-reported activity log data and objective tracking; yet, the mean absolute difference between tracking methods was 96 minutes (SD = 8.6) and suggests over- or under-reporting of families. Children used the modified ride-on car more in the first half compared to the second half of the three-month period (p < .05). Conclusions: This study may inform future research studies and local chapters of the Go Baby Go community program.

As robotics technology improves, remotely-operated telepresence robots will become more prevalent in homes and businesses, allowing guests, business partners, and contractors to visit and accomplish tasks without being physically present. These devices raise new privacy concerns: a telepresence robot may be used by a remote operator to spy on the local area, or recorded video may be viewed by a third party. Video filtering is one method of reducing spying ability while still allowing the remote operator to perform their task. In this paper, we examine the effects of three different visual conditions (filters) on the remote operator’s ability to discern details while completing a navigation task. We found that applying such filters protected privacy without significantly affecting the operator’s ability to perform the task, and that a depth image filter was the most effective privacy protector. We also found that the cognitive load of driving the robot has a slight privacy-protecting effect.

In this paper we propose a framework for conceptualizing and demonstrating a good-faith effort when developing autonomous systems. The framework addresses two fundamental problems facing autonomous systems: (1) the disconnect between human-mental models and machine-based sensors and algorithms; and (2) unpredictability in complex systems. We address these problems using a mix of education - explicitly delineating the mapping between human concepts and their machine equivalents in a structured manner - and data sampling with expected ranges as a testing mechanism.

In recent years, deep learning approaches have been leveraged to achieve impressive results in object recognition. However, such techniques are problematic in real world robotics applications because of the burden of collecting and labeling training images. We present a framework by which we can direct a robot to acquire domain-relevant data with little human effort. This framework is situated in a lifelong learning paradigm by which the robot can be more intelligent about how it collects and stores data over time. By iteratively training only on image views that increase classifier performance, our approach is able to collect representative views of objects with fewer data requirements for longterm storage of datasets. We show that our approach for acquiring domain-relevant data leads to a significant improvement in classification performance on in-domain objects compared to using available pre-constructed datasets. Additionally, our iterative view sampling method is able to find a good balance between classifier performance and data storage constraints.

Research into smart wheelchairs has been conducted for decades, but we have yet to see the widespread use of this technology among full-time wheelchair users. We argue that the main reason for this is that there is a mismatch between research and the actualities of using a powered mobility device in the real world. Based on our own research experiences, we enumerate some of these disparities, and offer some suggestions for where work in smart wheelchairs might focus in the coming years.

Research into smart wheelchairs has been conducted for decades, but we have yet to see the widespread use of this technology among full-time wheelchair users. We argue that the main reason for this is that there is a mismatch between research and the actualities of using a powered mobility device in the real world. Based on our own research experiences, we enumerate some of these disparities, and offer some suggestions for where work in smart wheelchairs might focus in the coming years.