Innovation Initiative “Project Linchpin”: Paving the Way for Army Artificial Intelligence

TL;DR:

  • The U.S. Army is advancing its AI solutions pipeline for sensing, targeting, and ISR operations.
  • Project Linchpin aims to expand the use of AI in tactical environments while ensuring data security and modularity.
  • Army leaders briefed industry and Army attendees about Project Linchpin during a recent meeting.
  • Contracts for the AI/machine learning pipeline are expected to be awarded in 2024.
  • The Army plans to issue a request for white papers and hold an Industry Day.
  • They are developing an ecosystem to handle “dirty” data and use multiple models for different problem sets.
  • Office space and personnel are being added to support the project.
  • The Army seeks to transform the vision of Project Linchpin into a long-term acquisition program.
  • The goal is to deliver “trusted AI” initially for IEW&S and potentially expand it to the entire Army.
  • The acquisition cycle is approached differently, with continuous competition and incentives for small businesses.
  • The solicitation will follow a phased contracting approach, starting with a standalone OTA.
  • AI models are considered disposable, and the Army aims to “rip and replace” algorithms.
  • Specific and innovative solutions from vendors are encouraged to accelerate progress in AI implementation.

Main AI News:

The U.S. Army is making significant strides in its pursuit of constructing a cutting-edge artificial intelligence (AI) solutions pipeline for its intelligence, surveillance, and reconnaissance (ISR) operations. Project Linchpin, a venture introduced last autumn by the Program Executive Office for Intelligence, Electronic Warfare and Surveillance (PEO IEW&S), aims to expand the utilization of AI in tactical environments while upholding data security, transparency, and modular AI systems that offer unparalleled flexibility, as explained by Army leaders.

At a recent Technical Exchange Meeting—TEM-X—in Philadelphia held on May 24-25, attended by 1,200 industry experts and Army personnel, Young Bang, Principal Deputy Assistant Secretary of the Army for Acquisition, Logistics and Technology (ASA ALT), Mark Kitz, Program Executive Officer of IEW&S, Alex Miller, Senior Advisor for Science and Technology and Vice Chief of Staff of the Army, and Colonel Christopher Anderson, Project Linchpin Program Manager for Intelligence Systems and Analytics at PEO IEW&S, provided a comprehensive briefing.

The Army’s intention is to award contracts for the AI/machine learning operations pipeline in March or April of 2024, pending the decisional guidance from senior leaders in November 2023. Given the size and strength of their force and soldiers, the Army leaders anticipate becoming the primary consumers of artificial intelligence and machine learning among the services. Young Bang, a former Army captain and West Point graduate who now specializes in AI, stated, “We are going to be the biggest consumers of artificial intelligence and machine learning in the services, given the density of our force and soldiers.”

To commence the process, the group plans to issue a request for white papers in September or October and organize a Project Linchpin Industry Day in August or September. The Army is actively working on drafting industry specifications by June or July to prepare for this event. Mark Kitz acknowledges the challenge of dealing with datasets that are inherently “dirty” and emphasizes the need to establish an ecosystem that embraces the fact that different models will yield varied performance across the diverse datasets available.

Col. Anderson reported that the Army is in the process of expanding office space and personnel for the project. He expressed gratitude for the support received from Army leaders in establishing what is expected to be a long-term acquisition program. The primary objective for Col. Anderson as the project manager is to transform this vision into a sustainable program of record. He stated, “My problem statement as PM [project manager] is how do you turn that vision into a program of record.” The origins of Project Linchpin lie in the Army’s recognition of the need for a data ecosystem that can serve as a foundation for model developers. This ecosystem will facilitate the evaluation of model effectiveness against relevant data, initially focusing on sensor data for IEW&S.

Despite initially targeting IEW&S, Principal Deputy Bang envisions that Project Linchpin’s artificial intelligence pipeline could expand to serve the larger Army. The leaders emphasized their unique approach to acquisition for Project Linchpin, one that deviates from traditional practices. They aim to create an environment of continuous competition within an industry partner ecosystem while incentivizing the utilization of small business solutions. Collaboration with ASA (ALT) Douglas Bush is underway to incorporate specific language in the upcoming request for proposals that promote the involvement of small businesses.

To achieve their AI-related objectives, the Army will adopt a phased contracting approach for solicitation. The first contract award, anticipated for next spring, will likely be a standalone other transaction authority (OTA) designed to expedite the development of prototypes and target AI-related technology services, products, and solutions. Following the OTA, the Army plans to offer further contract opportunities.

During the proceedings, Miller warned industry experts that the Army considers AI models as “disposable” and will avoid locking into a single solution or algorithm provided by any industry. He expressed a desire to have the ability to “rip and replace” algorithms as needed. Miller acknowledged the recent advancements in AI, such as ChatGPT, but cautioned that these capabilities are far from fulfilling the service’s requirements. He explained, “These generative AI models generate the exact response that we already know… That’s not telling us anything new. They are telling us what is statistically probable to be right.

Principal Deputy Bang encouraged industry representatives to offer specific and innovative solutions. While recognizing the vast scope of AI, encompassing pipelines, data management, training, and deployment, he emphasized that vendors claiming to cover all aspects would be dismissed. Instead, he welcomed vendors with expertise in specific areas, highlighting their potential to accelerate progress in other domains and facilitate the implementation of machine learning on a large scale.

As Project Linchpin forges ahead, the U.S. Army is at the forefront of harnessing the power of artificial intelligence and machine learning. With a well-defined vision and a strategic approach to acquisition, the Army aims to build an AI solutions pipeline that will revolutionize its sensing, targeting, and ISR operations, ultimately strengthening its capabilities and enhancing its mission success.

Conlcusion:

The U.S. Army’s ambitious pursuit of building an AI solutions pipeline through Project Linchpin signifies a significant development in the market. The Army’s emphasis on expanding the use of AI in tactical environments while ensuring data security and modularity presents a promising opportunity for technology companies and vendors specializing in AI and machine learning. The planned contracts, phased contracting approach, and focus on continuous competition and small business involvement indicate a dynamic market landscape with ample room for innovation and collaboration.

This initiative has the potential to stimulate the growth of AI-related products and services, as well as foster advancements in data management, algorithm development, and deployment solutions. Companies that can provide specific and innovative AI solutions tailored to the Army’s requirements may find themselves at the forefront of this emerging market, poised to contribute to the Army’s vision of achieving trusted AI and machine learning at scale.

Source