Distributed Analytics and Information Science International Technology Alliance

DAIS–ITA

DIAS logo

Research objective

DAIS–ITA (International Technology Alliance in Distributed Analytics and Information Sciences) is a collaborative partnership between the US Army and the UK Ministry of Defence. It brings together researchers from US Army Research Laboratories (ARL) and the UK Defence Science & Technology Laboratory (Dstl) to work alongside a consortium of universities and industrial research laboratories in US and UK. The goal of the alliance is to foster collaborative fundamental research in both nations to enable secure, dynamic semantically aware distributed analytics for situational understanding in coalition operations.

The consortium is led by IBM, which has major research and development operations in both nations.

This research program completed in September 2021.

Consortium members

  • University of California at Los Angeles
  • University of Massachusetts at Amherst
  • Pennsylvania State University
  • Purdue University
  • Stanford University
  • Yale University
  • Cardiff University
  • Imperial College London
  • University of Southampton
  • University College London
  • Raytheon BBN Technologies
  • Airbus Group
  • BAE Systems
  • IBM Research

Technical areas

Dynamic, secure coalition information infrastructures

Multidisciplinary research to provide the fundamental underpinnings to enable distributed, dynamic, secure coalition communication/information infrastructures that support distributed analytics to derive situational understanding.

Coalition distributed analytics & situational understanding

Multidisciplinary research to provide the fundamental underpinnings for future coalition distributed analytics and situational understanding in the context of ad hoc coalition operations at the tactical edge.

Collaborative research projects

1 Policy Enabled Dynamic Infrastructure

Exploring the ability to rapidly compose coalition infrastructures to support distributed analytics with easy configuration, on-demand resource allocation and resilience.
Specific areas of research include infrastructure design and distributed control for dynamic software defined coalition and federated policy learning and management.

2 Federated Learning for Coalition Analytics

Leveraging analytics and AI insights available from other coalition partners through distributed online learning with multiple learners, agile analytics enabled by decentralized continuous learning in coalitions and cognitive workflows and goal directed distributed analytics using semantic vector spaces.

3 Defending Coalitions in Adversarial Environments

Having the ability to intelligently share insights in environments with incomplete trust.
Specific areas of research include interpretability of neural networks in distributed and contested environments under incomplete trust and network intelligence from negative ties.

4 Ad-hoc Coalition Teams

Understanding how coalition teams, processes and policies would impact operations through coherence in coalitions, learning and inferencing in neuro-symbolic hybrids and a neural-symbolic learning of generative policies in coalition environments.

 

Past projects

1 Software-defined coalitions

Exploring the principles by which different elements across a coalition could be composed via control plane interactions to form a virtualised larger element.

2 Generative policy models for coalitions

Investigating approaches for policy-based management in a coalition environment with sufficient autonomy to its constituent elements.

3 Agile composition for coalition environments

Exploring new architectures in which analytics code and data of various types (ISR, HUMINT etc) are mobile and composed optimally.

4 Instinctive analytics in a coalition environment

Exploring how information systems adapt to user context and exploiting a future heterogeneous edge compute paradigm.

5 Anticipatory situational understanding for coalitions

Researching verifiable predictive analytics operating synergistically between users and machine learning/reasoning.

6 Evolution of complex adaptive human systems

Understanding complex adaptive human groups, their mutability and evolution.

 

Publications

All publications from the DAIS–ITA research program can be found on the publically available Science Library with full details of the authors, projects and topics.