Laura Carlson - US grants

In everyday life we constantly talk about objects in the world around us: we may enlist help in looking for an object (e.g., `Where are my keys?`); we may point out a surprising object in the environment (e.g., `Look at that billboard in front of the building!`); or, we may give instructions to put objects in certain locations (e.g., `Put the lamp to the left of the chair.`). In each case we must perform a mapping between the linguistic elements in the utterance and our perception of the objects and their relation in the world. For example, the words `lamp` and `chair` need to be mapped onto their appropriate referents in the world, and the relation `to the left of` needs to be mapped onto the spatial relationship between these objects. This project focuses on how linguistic spatial terms are mapped onto spatial relations in the environment. Two specific aspects of the process will be examined. The first is how spatial terms are mapped onto directions in space. This is an interesting topic because sometimes more than one mapping is possible. For example, in the utterance `Put the lamp to the left of the chair,` `left` could refer to the speaker's left side or to the chair's left side. Because these disagree for a speaker facing a chair, there is some ambiguity as to where the lamp should be placed. Understanding a potentially ambiguous utterance may therefore be difficult if all possible interpretations are considered. Previous research suggests that multiple interpretations of spatial relation terms are initially considered. In the present project one series of experiments uses an eye tracking methodology to examine the simultaneous activation more closely, with the additional goal of examining how the degree of disagreement among the varying interpretations influences the degree of difficulty in understanding the utterance. Another series of experiments examines how one interpretation is selected when multiple interpretations are initially activated. The specific question is whether the selection of one interpretation is accompanied or assisted by an inhibition of the other interpretations. The second aspect of the mapping process examined in the present project involves dividing up space around the objects into regions to which the relation term (e.g., `above`) does and does not apply. A series of experiments explores how such regions are defined, and examines factors that may influence the size of these regions, such as the conceptual size of the objects, the presence of other objects in the scene, or the functional relationship between the two objects. Together these experiments should further our understanding of the online interpretation of spatial relations. This has important implications theoretically for our understanding of the interface between language and spatial representation, and practically for our ability to use effective and unambiguous communication.

When people communicate with each other about spatially oriented tasks, they more often use qualitative spatial references (such as "behind" in the spatial description "Your eyeglasses are behind the lamp.") rather than precise quantitative terms. Although natural for people, such qualitative references are problematic for robots that "think" in terms of mathematical expressions and numbers. Yet, providing robots with the ability to understand and communicate with these spatial references has great potential for creating a more natural interface mechanism for robot users. This would allow users to interact with a robot much as they would with another human, and is especially critical if robots are to provide assistive capabilities in unstructured environments occupied by people. This project will do the following: empirically capture and characterize the key components of spatial descriptions that indicate the location of a target object in a 3D immersive task embedded in an eldercare scenario; develop and refine algorithms that enable the robot to produce and comprehend descriptions containing these empirically determined key components within this scenario; and assess and validate the robot spatial language algorithm in virtual and physical environments. This project will train graduate students in an interdisciplinary setting that encompasses psychology, computer science and engineering), and will directly involve undergraduate students in the robotics work at Missouri and in the human subject experimentation work at Notre Dame. This project will lead to a better understanding of how robots can and should be used for this class of assistive tasks in an eldercare scenario.

The University of Notre Dame will undertake a project that addresses the question: Since cultivating cultures for ethical STEM requires effective leadership, what training content and methods most successfully turn students into ethical leaders? The literature tells us that ethical leaders are role models to their peers, colleagues, mentees, and students. The leadership programs at the University of Notre Dame will thus provide training in a manner that empowers the trainees to mentor others with the expectation that, as mentors, they will perform scientific and technological research with an eye to the implications for society and pass this approach on to their own students and peers. The research team will assess and compare two different leadership training programs, the first grounded in research on business and management ethics, the second in research on ethics training for STEM students. While the primary goal of these two lines of research is the same - cultivating ethical leaders - the theories behind them and methods used by trainers are often quite different. Both leadership programs build from previously gathered assessment data, methods, and tools, and are imbued with ethics; both train STEM graduate students using an immersive and cumulative service-learning model. The two educational programs evaluated by this project are train-the-trainer programs that have the power to create exponential change rather than simply incremental improvements - both at Notre Dame and beyond. A primary outcome of the project will be the creation of a blended program that combines the best features of each program and appeals to both the business and research ethics approaches. Through the compilation, analysis, and dissemination of this blended program broadly, we will be promoting the most successful content and methods that turn STEM students into ethical leaders, thereby cultivating cultures for ethical STEM.

Over three years, 93 STEM students will receive leadership training. Our research project will: (1) assess these two programs' effectiveness in training ethical leaders, and (2) by comparing the results across the programs, identify the training components that will best equip students to grapple with ethical research dilemmas, helping to establish an ethical lab and workplace culture. By assessing participants before, after, and one-year out from their respective programs, we will evaluate content retention, ethical decision-making, and the effectiveness of training methods and materials, and create an overarching program.

Traditional research ethics training is premised on the idea that a researcher is responsible only for the ethical conduct of their research. This project supplements that foundation to add consideration of the social consequences - and the ethical meaning - of introducing potentially disruptive innovations to society. Increasingly, the public expects scientists and engineers to take into account the potential social and ethical ramifications of the innovations they develop. Whether in industry or academia, it is no longer considered enough to focus narrowly on the technical breakthroughs and leave the rest of society to deal with the implications - implications that often become apparent only after the innovation has been deployed. But what might more thoughtful development approaches look like in practice? And can they be made practical? Work on the nascent concept of Responsible Innovation attempts to answer these questions. This project furthers this research by testing and innovating on existing concepts of RI while also providing training in Responsible Innovation to an interdisciplinary cohort of graduate students and postdocs. In turn, this cohort helps to guide a selection of actual emerging innovations as test cases.

For each of its three years, this project recruits a cohort of 9-12 graduate students and postdocs (organized into small teams of 3-4) drawn not only from the natural sciences and engineering, but also from the social sciences and humanities. In the first half of each year, these recruits are trained in existing frameworks for Responsible Innovation (RI) as well as the skills, like anticipation, communication, and leadership, which enable it. As they work through this literature from their different disciplinary perspectives, they also critique and contribute new ideas and insights to this emerging area, with each team arriving at their own working approach to RI. In the second half of each year, the trainees are connected to a real-world commercialization project through Notre Dame's technology transfer and incubation hub, the IDEA Center. The teams then implement their Responsible Innovation concept as consultants to an actual innovation project. This provides a valuable service to that project while also serving as a practical test of their ideas. In the final phase of each year, the teams write up their findings from this experience for publication, contributing new ideas and reporting their real-world experience with them. This approach allows four valuable goals to be pursued simultaneously: 1. Train students and researchers in the latest thinking on Responsible Innovation; 2. Put social scientists, humanists, engineers, and natural scientists into conversation with one another about the social and ethical impacts of science and technology research; 3. Contribute new approaches and refinements of Responsible Innovation by engaging a wider array of researchers who evaluate and test its real-world utility; 4. Help inform and guide a selection of local, real-world innovation projects.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.