U.S.-based drone companies AeroVironment and Dragoon, as well as U.S.-based software firms Swan and Auterion, were chosen to compete in the project called Artemis, DIU said in a news release. Notably, the two software companies are each partnering with separate unnamed Ukrainian drone manufacturers.

DIU initiated Artemis in response to a congressional mandate, which directed operational testing of low-cost loitering munitions that can fly in electromagnetic contested environments and be deployed in large numbers. The unit wants to have a successful prototype by the end of fiscal 2025.

“We are excited about the non-traditional companies who are providing low-cost, adaptable, long-range, UAS platforms with the potential to maximize operational flexibility for the Joint force,” Trent Emeneker, DIU’s Artemis program manager and contractor, said in a statement. “This was the intent of Congress’ direction to rethink how to get capabilities to the warfighter at speed and scale that can deliver much faster than traditional Programs of Record.”

After releasing a solicitation in October 2024, DIU and the Office of the Under Secretary of Defense for Acquisition and Sustainment evaluated 165 proposals from vendors, held flight demonstrations and then down-selected to four industry teams, according to DIU.

With contracts in place, “the next step is meeting an aggressive testing and integration schedule to complete prototyping and demonstrate success by the end of May 2025,” DIU stated in a release.

The solicitation called for one-way, ground-launched drones from commercial vendors with an operational range of 50 to 300 kilometers or more. DIU wants Artemis prototypes that can carry a 10-plus kilogram payload more than 50 kilometers, and are “capable of supporting high-speed, low-altitude, beyond line of sight flight operations in [disrupted, disconnected, intermittent, and low-bandwidth] environments,” according to the RFP. Ideally, the organization would like the drones to be able to carry a 25-plus kilogram payload upwards of 300 kilometers.

Officials emphasized the Artemis program is directly linked to emerging trends on modern battlefields. Throughout Russia’s full-scale invasion of Ukraine, inexpensive kamikaze drones from commercial vendors have provided warfighters on both sides with key capabilities. In the Middle East, Iranian-backed Houthis launched multiple complex attacks on U.S. Naval forces stationed in the Red Sea last year, as well.

“With Artemis, DIU and A&S are moving rapidly to provide an option for Services and Combatant Commands to choose from, delivered years in advance of current Program of Record timeframes,” DIU stated in a release.

The post DIU taps 4 vendors — including Ukrainian firms — for long-range kamikaze drones appeared first on DefenseScoop.

Known as Project Athena, the pilot aims to evaluate the use cases and cost models of commercially available genAI tech that can be used to support the service’s back-end office work. The effort is being led by the Army’s Chief Information Officer Leonel Garciga alongside the Office of Enterprise Management (OEM). The assessment is slated to end in April, after which the department hopes to create a list of capabilities that can be purchased by various service components based on their needs and mission requirements.

“We’re going to assess different tools so that we can equip Army organizations with information. What capabilities should you consider based on your use cases? What is the cost model and what do you need to know about that?” Jennifer Swanson, deputy assistant secretary of the Army for data, engineering and software, said Tuesday during a roundtable with reporters.

The goal for Project Athena isn’t to choose a single generative AI platform and mandate its use across the service, but rather provide options for Army offices that detail the pros and cons of each capability — including their different features, use cases, cost models and deployment architectures, Swanson said.

Over the last year, the Army and others at the Pentagon have worked to understand how emerging genAI platforms that leverage large language models (LLMs) can be integrated into the department.

While some efforts have looked into the technology’s applicability for warfighting functions, the most immediately promising use cases are those that support daily business operations. In October, the service announced a new pilot dubbed #CalibrateAI focused on simplifying repetitive and arduous tasks, that has since been brought under Project Athena.

“We really wanted to focus on where we had this opportunity to employ capability at scale, get some of those use cases operating [and] look at some different models,” Garciga said during the roundtable. He added that Project Athena is evaluating a range of tools — from commercial-off-the-shelf software that can be deployed on existing environments to niche, integrated LLMs for existing capabilities.

Garciga noted that generative AI has been very useful in supporting the Army’s legal teams, public affairs offices and recruiting efforts. The technology has also shown promise in assessing documents related to requests for information (RFI) and sifting through the Pentagon’s vast inventory of policy documents, Swanson added.

Because funding for the genAI tools will come from individual Army organizations that choose to purchase them, a large part of Project Athena has been dedicated to informing leaders about the actual cost of implementing the capabilities — which can require additional cloud compute and storage infrastructure that might become too expensive for some offices to manage. 

“We want to make sure that we’re informing them from the standpoint of, this is really what you need to consider when you’re spending that money to make sure that we’re getting the best deal for the Army, and to make sure that we are aware of all the bills that may come with tools,” Swanson said.

Some genAI tools and platforms are already operationalized, accredited and on the network through Project Athena, according to Garciga. Once the pilot concludes in April, the Army plans to publish guidance based on lessons learned that were documented through the effort and work on what the service needs to do at the enterprise level moving forward.

“We want to throw stuff on the network and just operationalize it, but a lot of this has also been, what does this mean from an enterprise perspective? How do I hook up identities to it? How do we work on where we put the data?” Garciga said. “We could get that a little bit standardized so it makes sense.”

The post Army evaluating generative AI tools to support business ops appeared first on DefenseScoop.

The effort, dubbed #CalibrateAI, comes on the heels of other efforts — such as #DefendAI, #BreakAI and #CounterAI — that have been launched as the service pursues a 500-day implementation plan for adopting these types of technologies and overcoming some of the challenges associated with them.

“Gen AI models present unique and exciting opportunities for the Army. These models have the potential to transform mission processes by automating and executing certain tasks with unprecedented speed and efficiency. Commanders and senior leaders should encourage the use of Gen AI tools for their appropriate use cases,” Army CIO Leonel Garciga said in a directive issued to the department in June.

Generative AI systems can create new content — such as text, images, audio and video — based on the data they’ve been trained on and human prompts. ChatGPT, a popular large language model, is one prominent example.

While a lot of focus in the defense community is on how artificial intelligence could aid and enable U.S. military operations on the battlefield, officials are also keen on employing the technology for back-office functions and to handle some of the drudge work that humans have previously had to perform.

On Friday, Jennifer Swanson, deputy assistant secretary of the Army for data, engineering and software, announced the launch of #CalibrateAI, which will “explore innovative applications” of the tech for Army acquisition activities.

The initiative will use a “cutting-edge” capability developed by the non-profit LMI at no cost to the Army, according to officials.

The tool, which is intended to “simplify repetitive and time-consuming tasks,” will use data analytics, machine learning and natural language processing to “deliver tailored responses that are highly relevant to specific topics, improving the efficiency and effectiveness of information retrieval and analysis,” the Army stated in a press release.

However, officials are also concerned about ensuring the security and accuracy of the Defense Department’s data and the outputs of gen AI tools.

In his June memo, Garciga noted that these technologies present unique challenges with regard to data privacy, security and control over the generated content. “Therefore, their use should be carefully evaluated and monitored,” he wrote.

The #CalibrateAI capability will be deployed in an Impact Level 5 (IL5) secure cloud environment and be able to handle controlled unclassified information (CUI) data. It includes customizable user-access controls to protect “need-to-know” info, per the release.

The Army is worried about potential hallucinations, a trend seen in other generative artificial intelligence systems that have been launched in the commercial sector. For example, large language models sometimes yield responses to user inputs that are factually inaccurate or otherwise problematic.

Recognizing these potential pitfalls, the new pilot will also focus on “identifying and calling out potential ‘hallucinations’ or erroneous outputs, thereby increasing the reliability of AI-generated content,” according to the release.

“By using off-the-shelf AI tools and leveraging cross-service authority-to-operate reciprocity granted by DoD CIO, #CalibrateAI will explore how we increase productivity while enhancing the accuracy of information,” Swanson said in a statement. “The ability to query curated document sets for generating new content, along with providing citations, will ensure that our outputs are not only accurate but also easily fact-checked.”

The broad goals of the pilot for the acquisition community include increasing productivity, improving accuracy, promoting innovation, and identifying a cost-effective route for wider adoption of the technology. According to officials, that will be achieved by using AI tools to collate, curate and generate critical information relevant to acquisition activities, implementing mechanisms to provide citations, and promoting the exploration of novel applications of artificial intelligence in acquisition.

The post Army kicks off generative AI pilot to tackle drudge work, hallucinations appeared first on DefenseScoop.

Officials published a new directive in March that aims to streamline and standardize the Army’s approach to developing software — an extensive policy outlining how it plans to develop requirements, perform testing, procure, sustain and manage personnel for software-based capabilities. The Army later posted a notice to Sam.gov in May which sought industry feedback on a multiple-award contract vehicle for its software efforts, estimated to be worth more than $1 billion over a 10-year period of performance.

Not long after, members of industry — mainly non-traditional and venture-backed defense contractors — began publicly sharing their concerns with both the directive and the request for information (RFI). Many took issue with the Army’s push to use cost-plus contracts for software “to the maximum extent possible” while minimizing firm fixed-priced models, as stated in the service’s software directive.

Several company executives, who all spoke to DefenseScoop on the condition of anonymity to talk candidly, claimed that the contract language showed preference for large, traditional defense prime contractors and would cause smaller, non-traditional vendors to avoid bidding on future Army software programs.

“It’s not worth trying to bid or convince them if what they’re stipulating as sort of a requirement to even enter isn’t something you can meet,” one defense tech executive said. “I think they would see a lot less innovation in their responses, and it sort of starts leading down the path of custom-build, services-companies approach — which in some cases could be appropriate, but we find it a little affronting for the Army to say, ‘You must agree to this as a condition to even participate.’”

Cost-plus versus firm-fixed-price

During a roundtable with reporters Sept. 5, Army acquisition lead Doug Bush emphasized that the service never planned to exclusively use cost-plus contract types for its future software programs. Rather, he said the organization is committed to customizing acquisition strategy to each specific capability and is open to using hybrid-contract approaches.

“Every contract needs to be tailored to the task, and it’s not a simple world of cost-plus versus fixed-price,” Bush said. “Those are ends of a spectrum. In between those things are a vast array of hybrid approaches that are tailored to different things.”

Under cost-plus contracts, the Defense Department agrees to pay contractors for a project’s expenses — including labor hours, materials and any other costs accrued while under the contract’s terms — as well as an additional fee to account for profit once the project is completed.

While cost-plus contracts do guarantee reimbursement for contractors and mitigate risk to them, the model requires companies to use government-approved cost-accounting systems that track their software engineers’ hours and other expenses for the contract. One defense tech executive explained that implementing and approving the necessary infrastructure is a time-consuming, expensive effort — especially for smaller companies.

Others shared that overall, cost-plus contracts have made development cumbersome, are contradictory to the culture of how non-traditional defense companies operate, and do not value software the way that it should be.

“Through a cost-type contract, it is valued through what the government can count,” a source told DefenseScoop. “It can count the number of people working on it, it can count the number of hours that they’re billing — and that is almost antithetical to how value is generated in software.”

On the other hand, many software-focused companies prefer contracting using a firm-fixed-price model, in which the Defense Department and vendor agree on a set price for products at the start of the contract. This means companies do take on added risk if there are cost overruns during the program, but they would not have to implement additional infrastructure to track labor hours.

“I think startups and more innovative companies — because of our model and how we’re valued in the market — we’re willing to eat some of the costs to make the changes and deliver the best thing where necessary,” one defense tech executive explained.

When asked why the service intended to maximize use of cost-plus contracts for software, an Army spokesperson told DefenseScoop that agile software development and continuous delivery hinge on iterative requirements based on user feedback. Therefore, assessing the price of a software capability before awarding the contract award becomes “impossible,” they said.

The Army has since clarified that it won’t solely be using cost-plus contract types for its software capabilities. In response to industry feedback, the service has added all contract types to the IDIQ in order to provide flexibility, Jennifer Swanson, the Army’s deputy assistant secretary for data, engineering and software, told reporters at the roundtable.

In an upcoming RFI set to be released in October, the Army will also update how it differentiates labor-category costs that will allow the service to better understand how industry pays its software engineers, she said. Where the government historically uses education degrees to differentiate, industry has largely moved away from that practice, she added.

The Army is also looking to initiate a pilot program that would allow the service to not specify contract types upfront in a task order requirement for the IDIQ. Rather, vendors can pitch software solutions and a proposed contract type they believe would fit best, giving both the service and industry more flexibility, Swanson said.

“The reason we’re doing that is so that we are able to consider all solutions that are out there,” she said. “There may be a commercial solution we’re not aware of, and so this would allow a company to bring that forward with a firm-fixed-price contract type that would make sense for a commercial solution.”

Although the Army is now considering all contract types moving forward, the service will use cost-plus models in some cases, Bush said.

“There will be times when a cost-type contract of some flavor is the appropriate thing to use to protect the government’s interest. I’m not going to apologize for that,” he told reporters. “This is the American people’s money. Our contracting approaches are gauged to both achieve the capability, but also make sure funding is not wasted.”

Remaining concerns

In the software directive, the Army noted that “[c]ustomization to commercial software solutions will be minimized to limit risk to the government. Where appropriate, microservices will be used to add capabilities not present in commercial software solutions. Customization to commercial software should only proceed where potential cost and technical risks are understood and mitigated.”

Defense tech executives told DefenseScoop that they believed the policy would steer the Army away from commercial products and services in favor of custom-built solutions — shutting themselves off to large segments of industry. Many non-traditional vendors have existing software products sold commercially that can be customized with new tools so they can be sold to the government as well, they said. 

“You can think about bringing into an architecture some software that was commercially developed that you are accessing by virtue of a license or some other type of agreement — a fixed-price contract agreement of some kind — because it’s already built, you only have to modify it,” one defense tech executive said.

The Army spokesperson said the service still fully supports customizing commercial software, but previous experiences in doing so have resulted in cost increases and cybersecurity risks to their systems, while also decreasing the capability’s quality and ability for the Army to update to newer versions. 

Because of those experiences, the service plans to “maximize configuration of commercial software while limiting customization” for systems using commercial software products at their core — such as most of the Army’s business systems, they said.

“This will reduce costs, increase capability and encourage more widespread use of commercial software while allowing the Army to leverage firm-fixed-price contracts,” the spokesperson said. “Commercial-off-the-shelf software products will remain the Army’s primary choice. The Army will only develop software when no viable commercial capability is available.”

One defense tech executive said that while they understood concerns with commercial software customization, industry must realize that not every software tool can meet the Army’s requirements — especially for military-specific capabilities.

“The easiest way is you use what is out there, but I think it’s incumbent upon industry to recognize that industry hasn’t created every piece of software that the government is going to need,” the defense tech executive said.

Despite some negative opinions, Bush emphasized that a majority of industry’s feedback to the RFI was positive. He said the Army continues to field industry’s responses and will reasonably address any concerns companies still have moving forward, adding that it’s okay for people to not fully agree with what the service is doing.

“Of course there’s always going to be some people who have concerns. And that’s fine,” he said. “I think we’ve listened to a lot of it. But if some companies don’t want to bid on a contract, it’s a free country. Don’t bid, others will. My goal is simply to get the capability for the Army, not to make everybody happy.”

A final request for proposals for the IDIQ contract vehicle will likely be published before the end of 2024, Swanson said.

The post Army continues refinement of software acquisition plans after industry criticism appeared first on DefenseScoop.

In May, the Army released a new directive to broadly adopt digital engineering techniques in order to improve weapon systems development — from the early acquisition stages of new capabilities to the sustainment of systems already fielded. Promoting interoperability and implementation throughout the department is one of four lines of effort outlined in the directive, and a critical hurdle will be ensuring that all of the digital engineering technologies the service uses are interoperable with one another.

To that end, the Army is supporting an update to the Systems Modeling Language standard that will be known as SysML 2.0 and is expected to be published before the end of the summer, said Jennifer Swanson, deputy assistant secretary for data, engineering and software. She told DefenseScoop during a roundtable Tuesday that the updated standard will serve as a near-term solution to the data-sharing challenge.

“There have not really been clear standards that digital engineering tools are developed to comply with, and so what we end up with is a lack of interoperability across these tools, which makes that data sharing between environments hard,” Swanson said.

SysML is an industry standard that provides general-purpose modeling language for systems engineering applications. Developed by a consortium of stakeholders from industry, government and academia, the standard supports system specification, analysis, design, verification and validation.

Swanson noted that by implementing the updated standards, vendors working with the Army will be able to share data across their digital environments more easily. Nearly all of the digital engineering vendors have communicated that they will adopt and implement SysML 2.0 “within the next year,” she said.

The inability to share data across multiple digital environments remains one of the top challenges to adopting digital engineering. That includes tools used by different industry partners, as well as those owned by the government, according to Swanson.

“You have the [program manager], who has established their own digital engineering environment using tools. There’s many, many tools on the market, so they’ve selected whatever tools they’ve picked. And then their vendors have digital engineering environments as well,” she said. “The Army is not going to direct everybody to use a certain tool, that doesn’t make sense for us to marry up with one specific vendor or technology.”

Although SysML 2.0 will be helpful for new programs that will leverage digital engineering from the beginning, Swanson noted the Army will also need to address data-sharing issues with programs already underway.

“We will have some challenges still with data that is in current digital engineering environments in terms of, now we have to figure out how to convert that to this new standard,” she said.

As it implements the directive, the Army is currently pulling in lessons from around 20 programs that are already using digital engineering tools as part of their development, such as the XM30 mechanized infantry combat vehicle program and the Future Long Range Assault Aircraft (FLRAA) effort.

The initiatives are two of six “pathfinder programs” identified by the service to act as guideposts for the Army’s broad implementation of digital engineering, each representing a different point of the development and sustainment cycle.

“We’re using these programs to help us take their lessons learned, what is working, what is not working, and be able to adapt all of that for more of an enterprise view across all of our programs to really help everybody else,” Swanson said.

For example, XM30 and FLRAA are both early in the acquisition process. Other pathfinders include the Joint Targeting Integrated Command and Control Suite (JTIC2S) — also in the early stages of acquisition — and the Integrated Fires Mission Command (IFMC), which is currently in production.

To inform sustainment efforts, the Army has chosen the M113 armored personnel carrier as its digital twins can apply broadly to the ground vehicle fleet, as well as legacy aircraft under the Program Executive Office Aviation Logistics Data Analysis Lab.

While these efforts already have experience in using digital engineering tools, Swanson noted they will still benefit from a robust adoption of the technology across the entire Army.

“The one thing that I think they are not benefiting from today is the other communities that we want to make sure we’re operating in these digital environments,” she said. “The requirements community, that test community, the sustainment community — we’re not there yet.”

The post New interoperability standards key to implementing Army’s digital engineering strategy appeared first on DefenseScoop.

LLMs, which can generate content — such as text, audio, code, images, videos and other types of media — based on prompts and data they are trained on, have exploded in popularity with the emergence of ChatGPT and other commercially available tools. Pentagon officials aim to leverage generative AI capabilities, but they want solutions that won’t expose sensitive information to unauthorized individuals. They also want technology that can be tailored to meet DOD’s unique needs.

“Definitely looking at pushing out guidance here, hopefully in the next two weeks, right — no promises right now, because it’s still in some staffing — on genAI and large language models,” Army CIO Leo Garciga said Thursday during a webinar hosted by AFCEA NOVA. “We continue to see the demand signal. And though [there is] lots of immaturity in this space, we’re working through what that looks like from a cyber perspective and how we’re going to treat that. So we’re gonna have some initial policy coming out.”

The CIO’s team has been consulting with the Office of the Assistant Secretary of the Army for Acquisition, Logistics and Technology as it fleshes things out.

“We’ve been working with our partners at ASAALT to kind of give some shaping out to industry and to the force so we can get a little bit more proactive in our experimentation and operationalization of that technology,” Garciga said.

Pentagon officials see generative AI as a tool that could be used across the department, from making back-office functions more efficient to aiding warfighters on the battlefield.

However, there are security concerns that need to be addressed.

“LLMs are awesome. They’re huge productivity boosters. They allow us to get a lot more work done. But they are very new technology … In my view, we are definitely in an AI bubble. Right? When you look kind of across industry, everybody’s competing to try to get their best, you know, LLM out there as quickly as possible. And by doing that, we have some gaps. I mean, we just do. And so it’s very important that we not take an LLM that is out on, you know, the web that I can just go and log into and access and put our data into it to try to get responses,” said Jennifer Swanson, deputy assistant secretary of the Army for data, engineering and software.

Doing so risks having the Army’s sensitive data bleed into the public domain via the internet and training models that adversaries could access, she noted.

“That’s really not OK, it’s very dangerous. And so we are looking at what we can do internally, within, you know, [Impact Level] 5, IL6, whatever boundaries, different boundaries that we have out there … And we’re moving as quickly as we can. And we definitely want tools within that space that our folks can use and our developers can use, but you know it’s not going to be the tools that are out there on the internet,” she added.

The forthcoming policy guidance is expected to address security concerns.

“I think folks are really concerned out in industry. And we’re getting a lot of feedback on, you know, just asking us what we think the guidance is going to look like. But we’re going to focus on putting some guardrails and some left and right limits … We’ve really focused on letting folks know, hey, this space is open for use. If you do have access to an LLM, right, make sure you’re putting the right data in there, make sure you understand what the left and right limits [are] … Don’t put, you know, an [operation order] inside public ChatGPT — probably not a good idea, right? Believe it or not, things like that are probably happening,” Garciga said.

“I think we really want to focus on making sure that it’s a data-to capability piece, and then add some depth for our vendors where we start putting a little bit of a box around, [if] I’m going to build a model for the U.S. government, what does it mean to for me to build it on prem in my corporate headquarters? What does that look like? … What is that relationship? Because that’s going to drive contracts and a bunch of other things. We’re going to start the initial wrapping of what that’s going to look like in our initial guidance memo so we can start having a more robust conversation in this space. But it’s really going to be focused around mostly data protection … and what we think the guardrails needs to be and what our interaction between the government and industry is going to look like in this space,” he added.

The post Army set to issue new policy guidance on use of large language models appeared first on DefenseScoop.

Gone are the days of stringent and onerous requirements. Now, for software intensive programs, industry can expect brief outlines of what the Army wants, Jennifer Swanson, deputy assistant secretary for data, engineering and software, told a forum.

“What we’re not doing and then not just moving away — we’re not doing these big, long monolithic requirements documents that lists every possible thing that one might ever want, things we don’t even know we want,” she said during the virtual Defense Software Modernization Forum hosted by ExecutiveBiz on Wednesday.

Going forward, “software CNS’s, capability needs statements, and software ICDs [initial capabilities documents], and those will be high level, brief, operationally focused [on] what is the capability that we’re looking for,” she said.

Then, the requirements owner and the operator the capability is intended for will guide developers during the software sprint, Swanson noted, adding they’ll refine the requirements on-the-fly to promote agility.

Industry has already seen changes with request-for-proposal language, Swanson said.

“Starting like last fall in our RFPs that are software heavy, where we are asking industry to ensure that they are using modern software practices and we’re going to be evaluating through orals and demos — not through written proposals, but through orals and demos as part of the evaluation process — not just the solution, but also the ability of the company to be agile,” she explained. “If we give you different requirements, show me that you can change it quickly, because that’s what we need. This is something I’ve been messaging to industry for really since last summer when we started because it’s a radical change in how we have approached software development to date in the Army. Clearly, we want industry to be able to be prepared for that.”

The draft needs statement and capability document are in staffing and nearly ready to go, Swanson said, calling this a “huge change.”

Recently, Swanson stressed the importance of software for the Army, noting that soldiers must be able to harness it at the tactical edge.

Software is an “enhancer to the force when they’re in a fight against a near-peer adversary, and we don’t really 100% know what they have or what’s going to come … What we’re trying to do is posture ourselves to be able to deliver those capabilities on-the-fly as needed, quickly,” she said at the Army’s technical exchange meeting in Philadelphia last month.

The post Army changing approach to software for new contracts appeared first on DefenseScoop.