Aaron Steinfeld

A robotic walker that provides guidance

This paper describes a robotic walker designed as an assistive device for frail elderly people with cognitive impairment. Locomotion is most often the primary form of exercise for the elderly, and devices that provide mobility assistance are critical for the health and well being of such individuals. Previous work on walkers focused primarily on safety but offered little or no assistance with navigation and global orientation. Our system provides these features in addition to the stability and support provided by conventional walkers. A software suite of robot localization and navigation combined with a shared-control haptic interface achieves this capability. The system has been tested in a retirement facility near Pittsburgh, PA, USA.

Field trial of Tiramisu: crowd-sourcing bus arrival times to spur co-design

Crowd-sourcing social computing systems represent a new material for HCI designers. However, these systems are difficult to work with and to prototype, because they require a critical mass of participants to investigate social behavior. Service design is an emerging research area that focuses on how customers co-produce the services that they use, and thus it appears to be a great domain to apply this new material. To investigate this relationship, we developed Tiramisu, a transit information system where commuters share GPS traces and submit problem reports. Tiramisu processes incoming traces and generates real-time arrival time predictions for buses. We conducted a field trial with 28 participants. In this paper we report on the results and reflect on the use of field trials to evaluate crowd-sourcing prototypes and on how crowd sourcing can generate co-production between citizens and public services.

Moving object detection with laser scanners

The detection and tracking of moving objects is an essential task in robotics. The CMU-RI Navlab group has developed such a system that uses a laser scanner as its primary sensor. We will describe our algorithm and its use in several applications. Our system worked successfully on indoor and outdoor platforms and with several different kinds and configurations of two-dimensional and three-dimensional laser scanners. The applications vary from collision warning systems, people classification, observing human tracks, and input to a dynamic planner. Several of these systems were evaluated in live field tests and shown to be robust and reliable.

The oz of wizard: simulating the human for interaction research

The Wizard of Oz experiment method has a long tradition of acceptance and use within the field of human-robot interaction. The community has traditionally downplayed the importance of interaction evaluations run with the inverse model: the human simulated to evaluate robot behavior, or “Oz of Wizard”. We argue that such studies play an important role in the field of human-robot interaction. We differentiate between methodologically rigorous human modeling and placeholder simulations using simplified human models. Guidelines are proposed for when Oz of Wizard results should be considered acceptable. This paper also describes a framework for describing the various permutations of Wizard and Oz states.

Impact of robot failures and feedback on real-time trust

Prior work in human trust of autonomous robots suggests the timing of reliability drops impact trust and control allocation strategies. However, trust is traditionally measured post-run, thereby masking the real-time changes in trust, reducing sensitivity to factors like inertia, and subjecting the measure to biases like the primacy-recency effect. Likewise, little is known on how feedback of robot confidence interacts in real-time with trust and control allocation strategies. An experiment to examine these issues showed trust loss due to early reliability drops is masked in traditional post-run measures, trust demonstrates inertia, and feedback alters allocation strategies independent of trust. The implications of specific findings on development of trust models and robot design are also discussed.

Interface lessons for fully and semi-autonomous mobile robots

Experts from the Robotics Institute were individually interviewed for their insight on interface lessons for fully and semi-autonomous mobile robots. Information was collected on four main themes: challenges, things that seem to work well, things that do not work well, and interface wisdom. The comments were then condensed and pooled into seven high-level categories: safety, remote awareness, control, command inputs, status and state, recovery, and interface design. Classification of expert comments was relatively straightforward in that many interviewees identified consistent material. This suggests that those producing interfaces for fully and semi-autonomous mobile robots should, at the minimum, ensure that they have addressed these broad topics.

Helping visually impaired users properly aim a camera

We evaluate three interaction modes to assist visually impaired users during the camera aiming process: speech, tone, and silent feedback. Our main assumption is that users are able to spatially localize what they want to photograph, and roughly aim the camera in the appropriate direction. Thus, small camera motions are sufficient for obtaining a good composition. Results in the context of documenting accessibility barriers related to public transportation show that audio feedback is valuable. Visually impaired users were not affected by audio feedback in terms of social comfort. Furthermore, we observed trends in favor of speech over tone, including higher ratings for ease of use. This study reinforces earlier work that suggests users who are blind or low vision find assisted photography appealing and useful.

Emergence of a Cognitive Car-Following Driver Model: Application to Rear-End Crashes with a Stopped Lead Vehicle

Rear-end crashes are a major roadway safety problem, and the potential of crash countermeasures to address this has long been recognized. High-frequency or severe-consequence scenarios are focused on the general lead-vehicle-not-moving (LVNM) case and specific crash scenarios. Operating scenarios are identified, and frequencies are assessed. From these, a small number of prevalent LVNM crash scenarios are identified as the focus for subsequent model development and crash counter-measure efforts. These scenarios suggest nominal atmospheric, roadway, lighting, vehicle, and driver conditions in designing cost-effective safety features to avoid LVNM rear-end crashes. From this, emergent models for cognitive car following are developed, based on fusing current knowledge. This will serve as a foundation for further model development efforts as well as for future human-factors experiments.

A user-centered drowsy-driver detection and warning system

This work is a culmination of years of research to develop an effective in-vehicle countermeasure to drowsy driving. Previous work resulted in an independently validated measure of drowsiness that was then incorporated into a drowsy-driver prototype monitor. The goal of this project was to develop an associated drowsy-driver interface that enabled effective, user-centered interactions with the underlying system.A multidisciplinary team designed a new drowsy-driver interface and introduced smart user interactions through a careful participatory design process that included both design experts and commercial motor vehicle drivers. It is hoped that this effort and subsequent field trials will result in a reliable, smart system that convinces drivers that they are driving in an unsafe condition and to make a wise choiceóstop and rest.

Robot confidence and trust alignment

Trust in automation plays a crucial role in human-robot interaction and usually varies during interactions. In scenarios of shared control, the ideal pattern is for the users real-time trust in the robot to align with robot performance. This should lead to an increased overall efficiency of the system by limiting under-trust and over-trust. However, users sometimes display incorrect trust and the ability to detect and alter user trust is important. This paper describes measures for real-time trust alignment.