skip to main content

Title: “Dave...I can assure you ...that it’s going to be all right ...” A Definition, Case for, and Survey of Algorithmic Assurances in Human-Autonomy Trust Relationships
People who design, use, and are affected by autonomous artificially intelligent agents want to be able to trust such agents—that is, to know that these agents will perform correctly, to understand the reasoning behind their actions, and to know how to use them appropriately. Many techniques have been devised to assess and influence human trust in artificially intelligent agents. However, these approaches are typically ad hoc and have not been formally related to each other or to formal trust models. This article presents a survey of algorithmic assurances , i.e., programmed components of agent operation that are expressly designed to calibrate user trust in artificially intelligent agents. Algorithmic assurances are first formally defined and classified from the perspective of formally modeled human-artificially intelligent agent trust relationships. Building on these definitions, a synthesis of research across communities such as machine learning, human-computer interaction, robotics, e-commerce, and others reveals that assurance algorithms naturally fall along a spectrum in terms of their impact on an agent’s core functionality, with seven notable classes ranging from integral assurances (which impact an agent’s core functionality) to supplemental assurances (which have no direct effect on agent performance). Common approaches within each of these classes are identified and discussed; benefits and drawbacks of different approaches are also investigated.  more » « less
Award ID(s):
Author(s) / Creator(s):
Date Published:
Journal Name:
ACM Computing Surveys
Page Range / eLocation ID:
1 to 37
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. The overall goal of our research is to develop a system of intelligent multimodal affective pedagogical agents that are effective for different types of learners (Adamo et al., 2021). While most of the research on pedagogical agents tends to focus on the cognitive aspects of online learning and instruction, this project explores the less-studied role of affective (or emotional) factors. We aim to design believable animated agents that can convey realistic, natural emotions through speech, facial expressions, and body gestures and that can react to the students’ detected emotional states with emotional intelligence. Within the context of this goal, the specific objective of the work reported in the paper was to examine the extent to which the agents’ facial micro-expressions affect students’ perception of the agents’ emotions and their naturalness. Micro-expressions are very brief facial expressions that occur when a person either deliberately or unconsciously conceals an emotion being felt (Ekman &Friesen, 1969). Our assumption is that if the animated agents display facial micro expressions in addition to macro expressions, they will convey higher expressive richness and naturalness to the viewer, as “the agents can possess two emotional streams, one based on interaction with the viewer and the other based on their own internal state, or situation” (Queiroz et al. 2014, p.2).The work reported in the paper involved two studies with human subjects. The objectives of the first study were to examine whether people can recognize micro-expressions (in isolation) in animated agents, and whether there are differences in recognition based on the agent’s visual style (e.g., stylized versus realistic). The objectives of the second study were to investigate whether people can recognize the animated agents’ micro-expressions when integrated with macro-expressions, the extent to which the presence of micro + macro-expressions affect the perceived expressivity and naturalness of the animated agents, the extent to which exaggerating the micro expressions, e.g. increasing the amplitude of the animated facial displacements affects emotion recognition and perceived agent naturalness and emotional expressivity, and whether there are differences based on the agent’s design characteristics. In the first study, 15 participants watched eight micro-expression animations representing four different emotions (happy, sad, fear, surprised). Four animations featured a stylized agent and four a realistic agent. For each animation, subjects were asked to identify the agent’s emotion conveyed by the micro-expression. In the second study, 234 participants watched three sets of eight animation clips (24 clips in total, 12 clips per agent). Four animations for each agent featured the character performing macro-expressions only, four animations for each agent featured the character performing macro- + micro-expressions without exaggeration, and four animations for each agent featured the agent performing macro + micro-expressions with exaggeration. Participants were asked to recognize the true emotion of the agent and rate the emotional expressivity ad naturalness of the agent in each clip using a 5-point Likert scale. We have collected all the data and completed the statistical analysis. Findings and discussion, implications for research and practice, and suggestions for future work will be reported in the full paper. ReferencesAdamo N., Benes, B., Mayer, R., Lei, X., Meyer, Z., &Lawson, A. (2021). Multimodal Affective Pedagogical Agents for Different Types of Learners. In: Russo D., Ahram T., Karwowski W., Di Bucchianico G., Taiar R. (eds) Intelligent Human Systems Integration 2021. IHSI 2021. Advances in Intelligent Systems and Computing, 1322. Springer, Cham., P., &Friesen, W. V. (1969, February). Nonverbal leakage and clues to deception. Psychiatry, 32(1), 88–106. Queiroz, R. B., Musse, S. R., &Badler, N. I. (2014). Investigating Macroexpressions and Microexpressions in Computer Graphics Animated Faces. Presence, 23(2), 191-208.

    more » « less
  2. In human-aware planning systems, a planning agent might need to explain its plan to a human user when that plan appears to be non-feasible or sub-optimal. A popular approach, called model reconciliation, has been proposed as a way to bring the model of the human user closer to the agent’s model. To do so, the agent provides an explanation that can be used to update the model of human such that the agent’s plan is feasible or optimal to the human user. Existing approaches to solve this problem have been based on automated planning methods and have been limited to classical planning problems only. In this paper, we approach the model reconciliation problem from a different perspective, that of knowledge representation and reasoning, and demonstrate that our approach can be applied not only to classical planning problems but also hybrid systems planning problems with durative actions and events/processes. In particular, we propose a logic-based framework for explanation generation, where given a knowledge base KBa (of an agent) and a knowledge base KBh (of a human user), each encoding their knowledge of a planning problem, and that KBa entails a query q (e.g., that a proposed plan of the agent is valid), the goal is to identify an explanation ε ⊆ KBa such that when it is used to update KBh, then the updated KBh also entails q. More specifically, we make the following contributions in this paper: (1) We formally define the notion of logic-based explanations in the context of model reconciliation problems; (2) We introduce a number of cost functions that can be used to reflect preferences between explanations; (3) We present algorithms to compute explanations for both classical planning and hybrid systems planning problems; and (4) We empirically evaluate their performance on such problems. Our empirical results demonstrate that, on classical planning problems, our approach is faster than the state of the art when the explanations are long or when the size of the knowledge base is small (e.g., the plans to be explained are short). They also demonstrate that our approach is efficient for hybrid systems planning problems. Finally, we evaluate the real-world efficacy of explanations generated by our algorithms through a controlled human user study, where we develop a proof-of-concept visualization system and use it as a medium for explanation communication. 
    more » « less
  3. The conventional machine learning (ML) and deep learning (DL) methods use large amount of data to construct desirable prediction models in a central fusion center for recognizing human activities. However, such model training encounters high communication costs and leads to privacy infringement. To address the issues of high communication overhead and privacy leakage, we employed a widely popular distributed ML technique called Federated Learning (FL) that generates a global model for predicting human activities by combining participated agents’ local knowledge. The state-of-the-art FL model fails to maintain acceptable accuracy when there is a large number of unreliable agents who can infuse false model, or, resource-constrained agents that fails to perform an assigned computational task within a given time window. We developed an FL model for predicting human activities by monitoring agent’s contributions towards model convergence and avoiding the unreliable and resource-constrained agents from training. We assign a score to each client when it joins in a network and the score is updated based on the agent’s activities during training. We consider three mobile robots as FL clients that are heterogeneous in terms of their resources such as processing capability, memory, bandwidth, battery-life and data volume. We consider heterogeneous mobile robots for understanding the effects of real-world FL setting in presence of resource-constrained agents. We consider an agent unreliable if it repeatedly gives slow response or infuses incorrect models during training. By disregarding the unreliable and weak agents, we carry-out the local training of the FL process on selected agents. If somehow, a weak agent is selected and started showing straggler issues, we leverage asynchronous FL mechanism that aggregate the local models whenever it receives a model update from the agents. Asynchronous FL eliminates the issue of waiting for a long time to receive model updates from the weak agents. To the end, we simulate how we can track the behavior of the agents through a reward-punishment scheme and present the influence of unreliable and resource-constrained agents in the FL process. We found that FL performs slightly worse than centralized models, if there is no unreliable and resource-constrained agent. However, as the number of malicious and straggler clients increases, our proposed model performs more effectively by identifying and avoiding those agents while recognizing human activities as compared to the stateof-the-art FL and ML approaches. 
    more » « less
  4. Enhancing resilience in distributed networks in the face of malicious agents is an important problem for which many key theoretical results and applications require further development and characterization. This work focuses on the problem of distributed optimization in multi-agent cyberphysical systems, where a legitimate agent’s dynamic is influenced both by the values it receives from potentially malicious neighboring agents, and by its own self-serving target function. We develop a new algorithmic and analytical framework to achieve resilience for the class of problems where stochastic values of trust between agents exist and can be exploited. In this case we show that convergence to the true global optimal point can be recovered, both in mean and almost surely, even in the presence of malicious agents. Furthermore, we provide expected convergence rate guarantees in the form of upper bounds on the expected squared distance to the optimal value. Finally, we present numerical results that validate the analytical convergence guarantees we present in this paper even when the malicious agents compose the majority of agents in the network. 
    more » « less
  5. null (Ed.)
    Conveying complex objectives to reinforcement learning (RL) agents can often be difficult, involving meticulous design of reward functions that are sufficiently informative yet easy enough to provide. Human-in-the-loop RL methods allow practitioners to instead interactively teach agents through tailored feedback; however, such approaches have been challenging to scale since human feedback is very expensive. In this work, we aim to make this process more sample- and feedback-efficient. We present an off-policy, interactive RL algorithm that capitalizes on the strengths of both feedback and off-policy learning. Specifically, we learn a reward model by actively querying a teacher’s preferences between two clips of behavior and use it to train an agent. To enable off-policy learning, we relabel all the agent’s past experience when its reward model changes. We additionally show that pre-training our agents with unsupervised exploration substantially increases the mileage of its queries. We demonstrate that our approach is capable of learning tasks of higher complexity than previously considered by human-in-the-loop methods, including a variety of locomotion and robotic manipulation skills. We also show that our method is able to utilize real-time human feedback to effectively prevent reward exploitation and learn new behaviors that are difficult to specify with standard reward functions. 
    more » « less