October 26, 2022 | International, C4ISR, Security
Harden the cybersecurity of US nuclear complex now
Buried in both the Senate and House versions of the 2023 National Defense Authorization Act are provisions augmenting the cybersecurity of the NC3.
September 4, 2018 | International, Land, C4ISR
New military drone roadmap ambivalent on killer robots
By: Kelsey Atherton
Drones are everywhere in the Pentagon today. While unpeopled vehicles are most closely associated with the Air Force and targeted killing campaigns, remotely controlled robots are in every branch of the military and used across all combatant commands. The fiscal year 2018 defense authorization contained the largest budget for drones and robots across the services ever, a sign of just how much of modern warfare involves these machines.
Which is perhaps why, when the Department of Defense released its latest roadmap for unmanned systems, the map came in at a punchy 60 pages, far shy of the 160-page tome released in 2013. This is a document less about a military imagining a future of flying robots and more about managing a present that includes them.
The normalization of battlefield robots
Promised since at least spring 2017, the new roadmap focuses on interoperability, autonomy, network security and human-machine collaboration.
The future of drones, and of unpeopled ground vehicles or water vehicles, is as tools that anyone can use, that can do most of what is asked of them on their own, that communicate without giving away the information they are sharing, and that will work to make the humans using the machines function as more-than-human.
This is about a normalization of battlefield robots, the same way that mechanized warfare moved from a theoretical approach to the standard style of fighting by nations a few generations ago. Network security isn't as flashy a highlight as “unprecedented battlefield surveillance by flying robot,” but it's part of making sure that those flying cameras don't, say, transmit easily intercepted data over an open channel.
“Future warfare will hinge on critical and efficient interactions between war-fighting systems,” states the roadmap. “This interoperable foundation will transmit timely information between information gatherers, decision makers, planners and war fighters.”
A network is nothing without its nodes, and the nodes that need to be interoperable here are a vast web of sensors and weapons, distributed among people and machines, that will have to work in concert in order to be worth the networking at all. The very nature of war trends toward pulling apart networks, toward isolation. Those nodes each become a point at which a network can be broken, unless they are redundant or autonomous.
Where will the lethal decision lie?
Nestled in the section on autonomy, the other signpost feature of the Pentagon's roadmap, is a small chart about the way forward. In that chart is a little box labeled “weaponization,” and in that box it says the near-term goals are DoD strategy assessment and lethal autonomous weapon systems assessment.
Lethal autonomous weapon systems are of such international concern that there is a meeting of state dignitaries and humanitarian officials in Geneva happening at the exact moment this roadmap was released. That intergovernmental body is hoping to decide whether or not militaries will develop robots that can kill of their own volition, according to however they've been programmed.
The Pentagon, at least in the roadmap, seems content to wait for its own assessment and the verdict of the international community before developing thinking weapons. Hedging on this, the same chart lists “Armed Wingman/Teammate (Human decision to engage)” as the goal for somewhere between 2029 and 2042.
“Unmanned systems with integrated AI, acting as a wingman or teammate with lethal armament could perform the vast majority of the actions associated with target identification,tracking, threat prioritization, and post-attack assessment," reads the report.
"This level of automation will alleviate the human operator of task-level activities associated with the engagement of a target, allowing the operator to focus on the identified threat and the decision to engage.”
The roadmap sketches out a vision of future war that hands off many decisions to autonomous machines, everything from detection to targeting, then loops the lethal decision back to a human responsible for making the call on whether or not the robot should use its weapons on the targets it selected.
Humans as battlefield bot-shepards, guiding autonomous machines into combat and signing off on the exact attacks, is a possible future for robots in war, one that likely skirts within the boundaries of still-unsettled international law.
Like its predecessor, this drone roadmap is plotting a rough path through newly charted territory. While it leans heavily on the lessons of the present, the roadmap doesn't attempt to answer on its own the biggest questions of what robots will be doing on the battlefields of tomorrow. That is, fundamentally, a political question, and one that much of the American public itself doesn't yet have strong feelings about.
https://www.c4isrnet.com/unmanned/2018/08/31/new-military-drone-roadmap-ambivalent-on-killer-robots
On the same subject
-
-
October 21, 2024 | International, Aerospace
Boeing sells defence surveillance subsidiary to Thales
Boeing has finalised a deal with Thales Defense & Security for the sale of its small defence subsidiary, Digital Receiver Technology (DRT).
-
January 23, 2020 | International, Aerospace, C4ISR
The Air Force tested its Advanced Battle Management System. Here’s what worked, and what didn’t.
By: Valerie Insinna WASHINGTON — The first field test of the U.S. Air Force's experimental Advanced Battle Management System in December was a success, with about 26 out of 28 capabilities showing some semblance of functionality during a recent exercise, the service's acquisition chief said Tuesday. But the service will seek to be more ambitious during a second demonstration in April, which will focus on space and bring in elements from U.S. Space Command and U.S. Strategic Command, said Will Roper, the Air Force's assistant secretary for acquisition, technology and logistics. "I am thrilled to say that 26 out of 28 things work. That is too high of a success rate at this point in time, but I'll take it. We should be taking more risk than that,” he told reporters during a roundtable. The three-day test took place at Eglin Air Force Base, Florida, and involved a potential cruise missile attack on the United States simulated by QF-16 drones. Through the exercise, Air Force F-22 jets, Air Force and Navy F-35 fighters, the Navy destroyer Thomas Hudner, an Army unit equipped with the High Mobility Artillery Rocket System, as well as special operators shared data in real time in ways the services cannot currently do in an operational environment. What will ABMS eventually look like? That's still a mystery, even to the Air Force, which wants to test different solutions for connecting platforms, crunching data and sending it to other assets with the goal of eventually fielding what works and abandoning what doesn't. “We gave the team the goals of: Pull what you can together in three and a half months to see how far we can stretch, how quickly we could achieve something,” said Air Force chief architect Preston Dunlap, who manages the ABMS effort. “We were quite happy actually, even with 10 percent solutions.” Here's a rundown of some notable successes so far, as well as major failures: The F-35 and F-22 were able to stealthily exchange data. Despite the two jets having advanced “sensor fusion” capabilities, the Air Force's two most advanced fighters can't really talk to each other. The F-35 uses the Multifunction Advanced Data Link, or MADL, to securely share sensitive information with other F-35s, while the F-22 has its own data link, the Intra-Flight Data Link, or IFDL. Even using a non-stealthy connection to share information has its limitations: While the F-35 can both transmit and receive data via the Link 16, which meets NATO standards, the F-22 currently can only receive data. However, the first ABMS test showed hopeful signs for fifth-generation fighter communication. The demonstration involved radio systems — built by F-35 prime contractor Lockheed Martin as well as Northrop Grumman, which manufactures key structures and mission systems for the aircraft, including MADL, Dunlap said. The demo also included Honeywell-made antennas built to speak across both MADL and IFDL, he added. Those systems were integrated onto a ground based rig that “look[ed] like a big piece of hardware with radios on it,” according to Roper. Then, the F-35 and F-22 flew over the system, exchanging data by bouncing it back-and-forth from the ground-based radios, Dunlap said. He noted that the test verified that existing technology can be used to overcome three obstacles: translating the F-35's MADL to the F-22's IFDL; moving data across the different frequencies; and securing the communication. "It was really herculean,” Dunlap said. "[The contractors] were excited by the speed of the acquisition team to get the ball going." During the next ABMS demo in April, the Air Force plans to stretch the capability by putting the translation system inside the unmanned Kratos XQ-58 Valkyrie for flight-based testing. “I also challenged the team to expand the amount of information translated between the different platforms so they can take advantage of new information on the displays,” Dunlap said. An AC-130 gunship connected with SpaceX's Starlink constellation. Although Dunlap did not provide much detail on this element of the exercise, he confirmed that the AC-130 was able to pass data through the constellation of small, high-bandwidth commercial internet satellites. The Air Force has shown interest in connecting its platforms to commercial broadband satellites through its Global Lightning experiment. A demonstration with Starlink and the KC-135 tanker aircraft is in the works, and the service also plans to evaluate equipment from Iridium, OneWeb and L3Harris. The Air Force created a cloud-based application for command and control. Typically, the service performs command and control from air operations centers — physical buildings where analysts sit in front of computers with specialized software that provides data from multiple assets, Dunlap said. Changes to software don't necessarily happen automatically, and they may require assistance from information technology experts. In the ABMS exercise, the Air Force demonstrated a cloud-based battle management and situational awareness application for the first time. It used a “CloudOne” system to host data up to the secret level, which will be a formative system underlying ABMS, Dunlap said. Both Amazon and Microsoft are involved in standing up the CloudOne technology, but Roper said the Air Force could use the Joint Enterprise Defense Infrastructure contract vehicle for CloudOne if JEDI winner Microsoft provides better rates. The robot dogs were a swing and a miss. U.S. Special Operations Command brought the robots that are capable of augmenting surveillance operations to the ABMS field test, but operators couldn't figure out how to connect them with the other platforms involved in the exercise. “We had some robot dogs — apparently those exist — that can go and do patrol. We were never able to patch their feeds in,” Roper said. There's hope for cybernetic canines becoming part of ABMS in the future though. Roper added that the ABMS team would be welcome to try to integrate the robots in future exercises. https://www.c4isrnet.com/air/2020/01/22/the-us-air-force-tested-its-advanced-battle-management-system-heres-what-worked-and-what-didnt/