Page 28 - FCW, November/December 2019

Public Sector Innovations
A pilot program involving government and nongovernment organizations gathered information from four databases via application programming interfaces. It
used natural language processing to extract information and eligibility criteria from unstructured descriptions of clinical trials.
Users can be authenticated by VA, the Centers for Medicare and Medicaid Services or both to obtain access to their personal data and then have that data used to query the National Cancer Institute’s database to search for relevant clinical trials.
A working prototype for that matching process was developed this past summer, and it is an important goal. But Alterovitz and his team have even bigger ambitions for the data ecosystem. By working through the data architecture, collaboration incentives and privacy controls, they are building a framework that could support a wide range of health initiatives.
The goal is for the system to become more of an “honest broker and testing platform [with] standardized agreements for how to test these protocols,” he said.
AI: Increasing Operational Readiness on the Bradley Fighting Vehicle
Tank-automotive and Armaments Command, Army Materiel Command, U.S. Army
Groundbreaking technological developments take vision, faith and just the right amount of patience. Bringing artificial intelligence-driven predictive maintenance capabilities to the Army’s aging Bradley Fighting Vehicle fleet took all of that, and much of it came from Lt. Col. Matthew Johnson.
Johnson, project management officer for Bradley vehicles, took the leap after a technical snag caused a team to lose 90 percent of its data.
“He could have stopped the funding
and everything after that episode of bad engineering, but he believed in the program, and we kept going,” said Russ Goodrich, vice president of government operations at
Uptake. The Defense Innovation Unit chose the company to assist the Army in the effort. Two weeks later, Goodrich said, the team
used AI-driven technology to successfully predict that a turret would fail. The group continued to have wins, predicting towing failures for vehicles, monitoring weapons training operations and collecting valuable data along the way.
The Bradley must shoot and move while collecting fault data, Goodrich said, “so the more it moves, the more it breaks [and] the more we can fine-tune our algorithms.”
The analytics effort, which ended its first phase in September, is one of the Army’s first forays into predictive maintenance and is expected to lay the foundation for other vehicle platforms. It’s a living example of what Defense Department leaders often say they want: fail fast to develop capabilities faster.
“All these lessons learned will be incorporated into other assets,” Goodrich said.
AI: Collaboration Accelerator Framework for Execution
U.S. Air Force
Riley Repko serves in many roles, including strategic adviser to the Air Force secretary and chief of staff on innovation and modernization issues and the service’s senior adviser to the Defense Department’s Joint Artificial Intelligence Center. But he’s also known for being decades ahead of the current technology.
He was instrumental is launching the
AI Collaboration Accelerator Framework
for Execution, which is a bit like Google combined with Yelp for Air Force AI solutions. When an airman needs to understand what technology solutions exist, AI CAFE taps a global, AI-fueled database and uses machine learning algorithms to identify risks and investments in emerging technology — effectively linking a problem with possible solutions in the private and public sectors.
AI CAFE, which is an evolution of a
concept Repko has been percolating for almost two decades, helps users and organizations determine their readiness for AI solutions and how the applications work, while facilitating market research.
“The tech caught up to Riley’s brain,” quipped Jamie Dos Santos, chairman of the board at Cybraics.
Repko’s vision was to deliver a solution that can provide the best options, and he was driven by the belief that no single entity has all the answers, Dos Santos added. And Repko received funding, built an architecture and turned out a minimally viable product in 90 days.
It’s information sharing at its finest because it gives contracting officers
the ability to align contract needs with nontraditional companies and experts, Dos Santos said.
“We all need this information at our fingertips to fight current and ever-evolving threats,” she added. “It’s not just good
for the problem-solver — it’s good for the universities, the students, military research and development, and private-sector efforts. Everybody learns from that.”
AI in Grants Management
National Science Foundation
For the past 65 years, the National Science Foundation has been the financial engine behind much of the nation’s technological innovations by funding the most promising research in a wide range of disciplines.
But funding requests are becoming more specialized and interdisciplinary, and NSF has struggled to identify outside experts who can evaluate such requests. Employees might have binders full of nuclear physicists on whom they can call, but they are less likely to know who could evaluate research into soil ecosystems, which requires expertise in chemical, biological and physical sciences along with data modeling, wireless communication, and cyber and physical systems.
To address that problem, a team of passionate volunteers from across the agency developed an artificial intelligence
26 November/December 2019 FCW.COM