Shiwali Mohan

I am an AI systems scientist at the Palo Alto Research Center. I study human-aware AI - how to design intelligent interactive systems that model, learn, and reason about their human collaborators. I am especially interested in methods for effective and robust human-agent collaborative behavior in real-world settings. My research advances methods for complex sequential decision making and leverages insights about human behavior from cognitive science, linguistics, and economics.

I have designed collaborative agents for a variety of application domains general purpose robots, preventive healthcare, sustainable transportation, and augmented reality learning. I am particularly motivated to build intelligent collaborative technology social good and public welfare. My work is interdisciplinary and has been published at venues for research on artificial intelligence (JAIR, AAAI, IAAI), human cognition (ICCM, ACS, BICA), human-machine interaction (ACM TiiS, IEEE RO-MAN) as well as in applications (JMIR, EMBC, ACM/AAAI AIES). My research has been supported by the US government (DARPA, AFOSR, ARPA-E, NSF) and commercial entities (Xerox Corporation, Kaiser Permanente). Currently, I am the principal investigator for PARC’s efforts on DARPA SAIL-ON and DARPA GAILA.

Humans of AI summarizes my ongoing research and outlines my research agenda on incorporating explicit models of humans in design of collaborative AI systems. My research statement explains it in more detail

I am a graduate of the Soar lab led by John E. Laird that traces its legacy to foundational research on human cognition by Allen Newell and human decision making by Herbert Simon. My graduate experience set me on this wonderful journey in studying the computational underpinnings of human intelligence and building intelligent systems that support people in their goals. PARC’s legacy in leveraging cognitive psychology to design more effective technology has only made this journey more adventurous.

I am active on sigmoid.social.

news

Mar 10, 2022	Giving an invited talk on cognitive science and collaborative robots at Human-Interactive Robot Learning at HRI 2022.
Mar 3, 2022	ITL research by Aaron Mininger and John Laird in the Soar group won the best demonstration award at AAAI 2022.
Jun 10, 2021	Research on natural human-robot teaching led by Preeti Ramaraj accepted for oral presentation at IEEE RO-MAN 2021
May 14, 2021	Giving a talk on cognitive science and robotics at Talking Robotics.
May 6, 2021	Looking forward to talking to graduate students at UM ECSEL+.

selected publications

IEEE RO-MAN

Unpacking Human Teachers’ Intentions For Natural Interactive Task Learning

Ramaraj, Preeti, Ortiz Jr, Charles L, and Mohan, Shiwali

30th IEEE International Conference on Robot and Human Interactive Communication 2021

Abs arXiv

Interactive Task Learning (ITL) is an emerging research agenda that studies the design of complex intelligent robots that can acquire new knowledge through natural human teacher-robot learner interactions. ITL methods are particularly useful for designing intelligent robots whose behavior can be adapted by humans collaborating with them. Various research communities are contributing methods for ITL and a large subset of this research is robot-centered with a focus on developing algorithms that can learn online, quickly. This paper studies the ITL problem from a human-centered perspective to provide guidance for robot design so that human teachers can interact with ITL robots naturally. In this paper, we present 1) a cognitive task analysis of an interactive teaching study (N=10) that extracts and classify various actions intended and executed by human teachers when teaching a robot; 2) in-depth discussion of the teaching approach employed by two participants to understand the need for personal adaptation to individual styles; and 3) requirements for ITL robot design based on our analyses informed by plan-based theories of dialogue, specifically SharedPlans.
ACM TiiS

Exploring the Role of Common Model of Cognition in Designing Adaptive Coaching Interactions for Health Behavior Change

Mohan, Shiwali

ACM Transactions on Interactive Intelligent Systems 2021

Abs arXiv HTML

Our research aims to develop intelligent collaborative agents that are human-aware - they can model, learn, and reason about their human partner’s physiological, cognitive, and affective states. In this paper, we study how adaptive coaching interactions can be designed to help people develop sustainable healthy behaviors. We leverage the common model of cognition - CMC [26] - as a framework for unifying several behavior change theories that are known to be useful in human-human coaching. We motivate a set of interactive system desiderata based on the CMC-based view of behavior change. Then, we propose PARCoach - an interactive system that addresses the desiderata. PARCoach helps a trainee pick a relevant health goal, set an implementation intention, and track their behavior. During this process, the trainee identifies a specific goal-directed behavior as well as the situational context in which they will perform it. PARCcoach uses this information to send notifications to the trainee, reminding them of their chosen behavior and the context. We report the results from a 4-week deployment with 60 participants. Our results support the CMC-based view of behavior change and demonstrate that the desiderata for proposed interactive system design is useful in producing behavior change.
ACM TiiS

Designing an AI Health Coach and Studying its Utility in Promoting Regular Aerobic Exercise

Mohan, Shiwali, Venkatakrishnan, Anusha, and Hartzler, Andrea

ACM Transactions on Interactive Intelligent Systems (TiiS) 2020

Abs arXiv HTML

Our research aims to develop interactive, social agents that can coach people to learn new tasks, skills, and habits. In this paper, we focus on coaching sedentary, overweight individuals (i.e., trainees) to exercise regularly. We employ adaptive goal setting in which the intelligent health coach generates, tracks, and revises personalized exercise goals for a trainee. The goals become incrementally more difficult as the trainee progresses through the training program. Our approach is model-based - the coach maintains a parameterized model of the trainee’s aerobic capability that drives its expectation of the trainee’s performance. The model is continually revised based on trainee-coach interactions. The coach is embodied in a smartphone application, NutriWalking, which serves as a medium for coach-trainee interaction. We adopt a task-centric evaluation approach for studying the utility of the proposed algorithm in promoting regular aerobic exercise. We show that our approach can adapt the trainee program not only to several trainees with different capabilities, but also to how a trainee’s capability improves as they begin to exercise more. Experts rate the goals selected by the coach better than other plausible goals, demonstrating that our approach is consistent with clinical recommendations. Further, in a 6-week observational study with sedentary participants, we show that the proposed approach helps increase exercise volume performed each week.
JAIR

Acceptable planning: Influencing Individual Behavior to Reduce Transportation Energy Expenditure of a City

Mohan, Shiwali, Rakha, Hesham, and Klenk, Matt

Journal of Artificial Intelligence Research 2019

Abs HTML

Our research aims at developing intelligent systems to reduce the transportation-related energy expenditure of a large city by influencing individual behavior. We introduce Copter - an intelligent travel assistant that evaluates multi-modal travel alternatives to find a plan that is acceptable to a person given their context and preferences. We propose a formulation for acceptable planning that brings together ideas from AI, machine learning, and economics. This formulation has been incorporated in Copter that produces acceptable plans in real-time. We adopt a novel empirical evaluation framework that combines human decision data with a high fidelity multi-modal transportation simulation to demonstrate a 4% energy reduction and 20% delay reduction in a realistic deployment scenario in Los Angeles, California, USA.
ACS

Characterizing an Analogical Concept Memory for Architectures Implementing the Common Model of Cognition

Mohan, Shiwali, Klenk, Matt, Shreve, Matthew, Evans, Kent, and Maxwell, John

In Proceedings of the Annual Conference on Advances in Cognitive Systems 2020

arXiv
AAAI

Learning Fast and Slow: Levels of Learning in General Autonomous Intelligent Agents.

Laird, John, and Mohan, Shiwali

In Proeedings of the AAAI Conference on Artificial Intelligence 2018