Heterogeneous Systems for Information-Variable Environments (HIVE)
Abstract
The US Department of Defense envisions the use of autonomy (independent robotic systems) in complex operational environments where there are substantial constraints on communications, sensing capability, and local processing. Many instances of autonomous systems exist that function very well in highly constrained situations and environments but fail to generalize to novel unconstrained environments. As such, foreseeable future autonomy will not have the ability to independently function in these scenarios. This report describes fundamental research into the issues underlying the control of such heterogeneous systems and aims to lay the groundwork for the development of generalizable methodologies for the effective integration of heterogeneous agents in dynamic, information-variable environments. Specifically, we focus on 4 research areas: 1) directly accounting for human variability to enable better integration of human decisions, 2) data fusion/computer vision (CV), 3) agent adaptation, 4) and networking dynamic teams of humans and machines. Additionally, we describe efforts to integrate these 4 research areas through both ongoing experiments and plans for future work. In the long term, these theories will enable the creation of an analytical framework for humanmachine network design and control and enable powerful, highly adaptive human autonomy systems that will be a major technological push affecting a broad range of applications.
Document Details
- Document Type
- Technical Report
- Publication Date
- May 01, 2017
- Accession Number
- AD1035246
Entities
People
- Allison Mathis
- Amar R. Marathe
- Benjamin T. Files
- Ethan Stump
- Garrett Warnell
- Heesung Kwon
- Hyungtae Lee
- Jonroy D. Canady
- Kim A Drnec
- William D. Nothwang
Organizations
- United States Army Research Laboratory