Back to news

March 29, 2019 | International, C4ISR

Palantir wins competition to build Army intelligence system

By Shane Harris

The Army has chosen Palantir Technologies to deploy a complex battlefield intelligence system for soldiers, according to Army documents, a significant boost for a company that has attracted a devoted following in national security circles but had struggled to win a major defense contract.

Industry experts said it marked the first time that the government had tapped a Silicon Valley software company, as opposed to a traditional military contractor, to lead a defense program of record, which has a dedicated line of funding from Congress. The contract is potentially worth more than $800 million.

The Army's decision to go with Palantir, which was co-founded by Peter Thiel, the billionaire investor and sometimes adviser to President Trump, brings to a close the latest chapter in a fierce competition.

In March 2018, the Army chose Palantir and Raytheon to vie for the next phase of the Distributed Common Ground System (or DCGS-A, for Army), which lets users gather and analyze information about enemy movements, terrain and weather to create detailed maps and reports in real-time. The system is designed to be used by soldiers fighting in remote, harsh environments.

But critics within the Army and in Congress have for years complained that DCGS-A cost too much and didn't deliver the intelligence and capabilities that soldiers needed. Some soldiers said the system was too hard to use and searched for alternatives.

Many became backers of Palantir, which sells to governments and businesses, including in the financial and health care sectors.

Palantir and its advocates argued that their software was cheaper and could meet all the Army's requirements. But Army brass defended their decision to pay for a custom-built platform.

In 2016, Palantir successfully argued in court that the government was required by law to consider purchasing commercial products, when available, rather than custom ones.

That sent the Army back to the drawing board and led to the face off between Palantir and Raytheon.

Before his death. Sen. John McCain (R-Ariz.) praised the new approach on Twitter, noting that after the Army had already spent $3 billion in development costs, “it was time to find another way.

Raytheon and Palantir were allowed to test their respective software platforms with a live audience of soldiers, who told them what they liked and didn't and what they would change. The two companies then refined their offerings to suit the Army's needs.

Traditionally, the government first chooses a company to build a system according to a set of detailed requirements. But this approach let the Army take both companies' products for a test drive before settling on the winner.

“The Army changed its approach to acquisition,” Doug Philippone, a former Army Ranger who leads Palantir's defense business, said in an interview.

He said the company was always confident it could win if it were allowed to adjust its technology after getting feedback from soldiers, who he said put the software through a rigorous test, even parachuting out of airplanes with reinforced laptops containing Palantir's software.

Chris Johnson, a spokesman for Raytheon, said the company was disappointed in the outcome. “We will wait for the Army's de-brief to understand their decision.”

The Army did not provide a comment for this story.

Raytheon and Palantir may compete for subsequent phases of work on the program.

Unlike most Silicon Valley start-ups, which aim to make their fortunes building consumer applications and software, Palantir at its founding set its sites on Washington, believing that its data analytics tools would find an eager market among U.S. spy agencies and the military, which are constantly trying to manage ever-expanding streams of information.

Philippone said the Army win had validated Palantir's strategy.

“We founded the company around solving this particular mission,” he said.

The company faced initial skepticism from investors, who thought it couldn't overcome entrenched bureaucratic interests and what they saw as political favoritism that led the Pentagon to spend billions every year with the same small group of Beltway contractors.

“Everyone told us we should stay away from Washington because it was corrupt and we didn't know how to play golf with senators,” Joe Lonsdale, a Palantir co-founder, said in a 2011 interview.

The company got an early investment in 2005 from In-Q-Tel, the CIA's venture capital arm, which tries to quickly develop technologies that the intelligence agency might use.

The In-Q-Tel connection helped Palantir get meetings with U.S. officials and intelligence analysts, and even test its software with the CIA's counterterrorism center, according to people familiar with the matter.

https://www.washingtonpost.com/world/national-security/palantir-wins-competition-to-build-army-intelligence-system/2019/03/26/c6d62bf0-3927-11e9-aaae-69364b2ed137_story.html

On the same subject

  • BAE Systems signs US Army deal for M777 gun structures

    January 6, 2024 | International, Aerospace

    BAE Systems signs US Army deal for M777 gun structures

  • GAO Chides DoD For Absence Of Cybersecurity Requirements

    June 8, 2020 | International, C4ISR, Security

    GAO Chides DoD For Absence Of Cybersecurity Requirements

    Overall, costs of major DoD acquisition programs have grown by 54 percent over their lifetimes and schedule delays average two years, GAO's annual report finds. By THERESA HITCHENS WASHINGTON: Five years after the Pentagon demanded every weapon system include the requirement that it be able to fight through Russian and Chinese cyber attacks expected on future battlefields, DoD “does not often include cybersecurity” in key performance parameters (KPP) for major programs, says GAO in its annual defense acquisition review. Of the three services, the Air Force is the worst at fulfilling two of the three best cybersecurity practices, the report says. The congressional watchdog found “inconsistent implementation of leading software practices and cybersecurity measures” among high-dollar “major defense acquisition programs” (MDAPs) — 85 programs worth $1.80 trillion at the end of 2019. “This included longer-than-expected delivery times for software and delays completing cybersecurity assessments— outcomes disruptive to DOD's efforts to keep pace with warfighters' needs for enhanced, software-dependent capabilities and protect weapon systems from increasingly sophisticated cybersecurity threats,” GAO said in the June 3 report. Cybersecurity KPPs Left Out The GAO report explains that KPP “are considered the most critical requirements by the sponsor military organization, while key system attributes (KSA) and other performance attributes are considered essential for an effective military capability.” In 2015, DOD modified its main requirements policy—the Joint Capabilities Integration and Development System Manual (JCIDS) rules on “survivability” requirements to include the ability to operate in a “degraded cyber environment.” Yet, GAO found that, at the end of 2019, 25 of the 42 major acquisition programs reviewed regarding cybersecurity practices failed to include cybersecurity as a parameter in their KPPs; “even more programs reported that their KSAs did not address cybersecurity.” GAO has targeted cybersecurity, software development and DoD-wide information technology (IT) improvement programs in its recent annual reviews because DoD weapon systems “are more networked than ever before — a change that while providing benefits for the warfighter also “has come at a cost” because “more weapon components can now be attacked using cybersecurity capabilities,” GAO explains. “Further, networks can be used as a pathway to attack other systems.” The watchdog has found consistently that failing to bake in cybersecurity requirements to system design and development ends up costing more money and time when program offices struggle to re-engineer systems once they hit production. This is a problem that affects most types of software development; and similarly trying to upgrade or replace software to improve cybersecurity often proves impossible. The 2019 report thus “looked at DOD's progress with developing: (1) strategies that help ensure that programs are planning for and documenting cybersecurity risk management efforts (cybersecurity strategies), (2) evaluations that allow testers to identify systems' weaknesses that are susceptible to cybersecurity attacks and that could potentially jeopardize mission execution (cybersecurity vulnerability evaluations), and (3) assessments that evaluate the ability of a unit equipped with a system to support assigned missions (cybersecurity assessments).” Most of the 38 MDAPs reviewed reported creation of cybersecurity strategies. However, of the 19 major programs that require cybersecurity vulnerability evaluations — under regulations set by the Office of the Undersecretary of Defense for Acquisition and Sustainment Ellen Lord — 11 have not completed them or failed to do so on time. Another three said they didn't have a schedule yet for doing so; and one — an unnamed Air Force program — told GAO it actually didn't know if it had undertaken the required evaluation. Indeed, the Air Force had the worst record on the evaluations, with none of its six programs having completed the evaluation processes. Of the 42 programs, 14 told GAO they had not finished their cybersecurity assessments. GAO also “found variation among the military departments in the rates they had completed these assessments. Specifically, among the three military departments, the Army reported the best rate for programs conducting cybersecurity assessments, while the Air Force had the lowest rate.” IT and Software Problems Plague Programs “Over the years, weapon acquisition program officials, through their responses to our questionnaires, have consistently acknowledged software development as a risk item in their efforts to develop and field capabilities to the warfighter, and this year is no different,” GAO reported somewhat wryly. GAO found that more than a quarter of the 42 MDAPs reviewed reported cost growth from software changes but admitted that “details are limited” in DoD reporting. Part of that uncertainty might be due to the fact that GAO found a number of major programs are transitioning to commercial approaches to software development, such as “agile development” that involves introducing incremental improvements over time. However, GAO found, “deliveries often lag behind industry standards.” Indeed, Air Force acquisition czar Will Roper told a webinar yesterday sponsored by Dcode, a tech innovation hub connecting commercial industry to government agencies, that while the Air Force can't go back and re-do old programs, “every new contract we do has to include DevSecOps.” “We are all in,” he added, “it's going to change the world.” DevSecOps stands for “development, security and operations,” and is a framework and tools for “designing in” software and cybersecurity. Roper long has been a key champion within DoD for moving to commercial practices and has repeatedly said he wants the Air Force to become a “software company.” GAO said that officials from 26 of the MDAPs regarding software development reported that software concerns had created risks at some point during their program's history. The biggest problem faced was — you guessed it — changes necessitated to ensure cybersecurity. The second biggest program was that the software development simply was “more difficult than expected.” Hardware design changes also played a big role in creating software problems, requiring subsequent changes in software configurations. Interestingly, while often bemoaned as a cause for program delays, requirements changes came in at the low of end of the reported issues troubling software development. Of the 15 major DoD IT programs reviewed, worth $15.1 million, 10 had delays in their original baseline schedules. But on the bright side, 11 showed decreased life cycle cost estimates. Further, all 15 have cybersecurity strategies as required by DoD regulations, and most reported having undertaken in 2019 at least one operational cybersecurity test. That said, “less than half reported conducting developmental cybersecurity testing,” GAO found. And according to DoD's own “Cybersecurity Testing and Evaluation Guidebook,” GAO scolds, “not conducting developmental cybersecurity testing puts programs at an increased risk of cost and schedule growth and poor program performance. Cost and Schedule Growth Stabilizes As it does every year, GAO also reviewed all 85 MDAPs for cost and schedule growth, and on that front the news is good: GAO found that the programs DoD Overview “have generally stabilized non-quantity related — (i.e. meaning not related to buy more stuff) — cost growth and schedule growth.” “Between 2018 and 2019, total acquisition cost estimates for DoD's 85 current MDAPs grew by a combined $64 billion (a 4 percent increase), growth that was driven by decisions to increase planned quantities of some weapon systems,” GAO found. “For example, DoD more than doubled in the past year the total number of missiles it plans to acquire through the Air Force's Joint Air-to-Surface Standoff Missile program.” And some programs actually lowered their year-average costs. GAO found that 55 MDAPs (more than half) “had lower average procurement unit costs since last year. Examples of programs with lower unit costs include the Navy's Joint Precision Approach and Landing System (16 percent decrease) and the Air Force's F-22 Increment 3.2B Modernization (15 percent decrease).” “Also between 2018 and 2019, capability delivery schedules for MDAPs increased, on average, by just over 1 month (a 1 percent increase),” GAO said. However, the report cautioned that cost/schedule performance looks “less encouraging as measured against their original approved program baselines.” The report found that the major acquisition programs “have accumulated over $628 billion (or 54%) in total cost growth since program start, most of which is unrelated to the increase in quantities purchased. Additionally, over the same time period, time required to deliver initial capabilities has increased by 30%, resulting in an average delay of more than two years. https://breakingdefense.com/2020/06/major-dod-acquisition-programs-flounder-on-cybersecurity-gao

  • The US Army is building a ‘cloud in the sky’ for its aviation fleet

    October 15, 2019 | International, Aerospace

    The US Army is building a ‘cloud in the sky’ for its aviation fleet

    By: Jen Judson WASHINGTON — The Army is building what is essentially a “cloud in the sky” for its current aviation fleet as it prepares the aircraft to fight alongside a future fleet under development, according to Brig. Gen. Thomas Todd, the program executive officer for Army aviation. The general spoke to Defense News in an interview ahead of the Association of the U.S. Army's annual conference. The current fleet won't dissolve into thin air when future helicopters are fielded, and they will be expected to fly together in operational environments across multiple domains. The Army is aiming to field a future attack reconnaissance aircraft and a future long-range assault aircraft by 2030. “We have to figure out a way to host a common server so that we can store data, process data and transport data quicker,” Todd said. “So while they work on the future vertical lift architecture, we still have to make the enduring fleet, that will fly alongside it, work and be capable.” So the Army is building an Aviation Mission Common Server, or AMCS, that is a stack of storage, data processing and transport capability “that's very much a flying cloud, if you will,” Todd said. The AMCS “will reside inside every aircraft. It has to reside in every aircraft because there has to be onboard processing and storage power,” Todd said, adding that it will be the engine that drives the associated user interface and apps as well as provide connectivity to the network overall. The user interface will be built upon the technology developed by Northrop Grumman for the Victor-model Black Hawk. The "V" model is an L-model UH-60 with a digital, modern cockpit like the "M" model, the latest Black Hawk variant, but not with an M-model price tag. Additionally, the interface in a V model can take on new capability through apps like a smartphone. The V model wrapped up its initial operational test and evaluation in September at Joint Base Lewis-McChord in Washington state, Brig. Gen. MacMcCurry, who is the Army G-3/5/7 aviation director inside the Pentagon, told Defense News in a separate interview. He reported the tests went well and the service looks forward to building out the fleet. The Army is currently assessing integrating the same user interface into Mike-model Black Hawks, according to Todd. The effort to build the server is part of a larger effort to ensure the current fleet is ready to fight in multidomain operations. The Army wants to obtain multidomain dominance by 2035. “We took a look at Army Futures Command's guidance on exactly what those combat aviation brigades would have in them and what would be enduring. For example, the Apache would be there indefinitely,” Todd said. “We also found that the requirements for data, the transport of and use of was exponential. So ultimately it's a problem that exists for the entire fleet, so we need to get after, at a minimum, making the enduring fleet compatible with future vertical lift, if not more capable.” There are several cross-cutting initiatives for the current fleet to make the aircraft more agile, interoperable, survivable and integrated in multidomain operations, Todd said, and the network will play an integral role. For example, the Army is working with the network community to replace its AN/ARC-201 radios with radios with the TSM waveform, which will improve and comply with future air-to-ground radio communications, according to Todd. There is also work being done within the position, navigation and timing community and with the Air Force to develop antennas, processors and software that hep the current fleet to survive battle, Todd added. And the service is working to improve power sources onboard aircraft. “There is a huge demand requirement coming, a demand signal for onboard systems and the power requirements of those is exponential. So given that it's not linear and it's going to grow exponentially, we have to get after alternative means,” Todd said. The same team that is in charge of the Improved Turbine Engine Program — which will replace engines in Apaches and Black Hawks and be the engine for the future attack reconnaissance aircraft — is looking at supplemental power units, upgrades to generators and upgrades to batteries to better power onboard systems that may not require the main power system anymore, Todd said. https://www.c4isrnet.com/2019/10/15/the-us-army-is-building-a-cloud-in-the-sky-for-its-aviation-fleet

All news