Keywords

Additional Keywords not in the list above

Learning Objective 1

Learning Objective 2

Learning Objective 3

Learning Objective 4

Abstract

Human-Machine Team (HMT) research represents one of the fastest growing fields of inquiry in science and technology. The need for HMT research is accelerated by warfare requirements, the AI revolution, and technological advances yielding a wide variety of collaborative automated platforms to support warfighters' and operators' decision-making in current and future warfare paradigms. However, there is a sizable confusion on what are the main factors and approaches for investigating HMTs to produce effective systems. Additionally, without referencing the vast existing knowledge of what makes human teams perform effectively and applying those lessons to HMTs, we risk reinventing the wheel, or worse, we risk neglecting human factors considerations, thereby leading to infamous “human error” outcomes and poor acceptance of HMT technologies.

The purpose of this tutorial is to provide HMT stakeholders, whether they be scientists, engineers, or decision-makers, with a practical guide to address HMT science and technology holistically. The goal of this tutorial is to present logically and simply the most salient aspects of HMT research for stakeholders to develop a robust understanding of HMT research and its desired impact on the development of HMT technologies that support and extend warfighters' capabilities.

The tutorial begins by providing contextual and historical background on HMTs, pointing to a rapid paradigm shift where perceptions of machines evolved from simple subordinates with precise tasking, to collaborative synthetic teammates supporting decision-making processes and autonomously carrying out mission objectives. We illustrate how this accelerated shift is driven by the AI revolution, future warfare demands, and modern mythology. Next, we provide an overview of how this shift also impacts Level of Automation (LOA) taxonomies in terms of requiring additional definition for collaborative human-machine decision-making, while providing practical examples.

Our focus turns to HMT research, and the need to align to DoD priorities. To that end, we first introduce the automation vision from the DoD Communities of Interest (COI). Second, we outline HMT research gaps and roadmaps from a seminal consensus study. And finally, we introduce how AI-driven automation needs to follow risk management best practices as well as DoD ethical principles for developing responsible AI.

Finally, the tutorial addresses important HMT performance enablers, focusing on three main enablers: calibrated trust, team situation awareness, and adaptive Human-Machine Interfaces (HMI). Our conclusion will summarize the tutorial's main points under the lens of conducting effective HMT research while providing useful resources to the practitioner in support of that endeavor.