September 19, 2018 | International, C4ISR

Differentiating a port from a shipyard is a new kind of problem for AI

It's well known that satellites and other intelligence, surveillance and reconnaissance platforms collect more data than is possible for humans to analyze.

To tackle this problem, the Intelligence Advanced Research Projects Activity, or IARPA, conducted the Functional Map of the World (fMoW) TopCoder challenge from July 2017 through February 2018, inviting researchers in industry and academia to develop deep learning algorithms capable of scanning and identifying different classes of objects in satellite imagery. IARPA curated a dataset of 1 million annotated, high-resolution satellite images aggregated using automated algorithms and crowd sourced images for competitors to train their algorithms to classify objects into 63 classes, such as airports, schools, oil wells, shipyards, or ports.

Researchers powered their deep learning algorithms by combining large neural networks, known as convolutional neural networks (CNNs), and computers with large amounts of processing power. The result was a network that, when fed massive amounts of training data, can learn to identify and classify various objects from satellite imagery. By combining a number of these networks into what is called an ensemble, the algorithm can judge the results from each CNN to produce a final, improved result that is more robust than any single CNN.

This is how a team from Lockheed Martin, led by Mark Pritt, designed their deep learning algorithm for the challenge. Pritt explained to C4ISRNET, that he and his team developed their CNN using machine learning software and framework from online open source software libraries, such as Tensor Flow. Earning a top five finish, the algorithm designed by Pritt's team achieved a total accuracy of 83 percent, and was able to classify 100 objects per second. Pritt said that with fully functioning algorithm, this software could take an image recognition task that takes a human an hour to complete and reduce the process to a few seconds.

The team's algorithm excelled at identifying classes with distinctive features, and successfully matched nuclear power plants, tunnel openings, runways, tool booths, and wind farms with accuracies greater than 95 percent, but struggled with more indiscreet classes such as shipyards and ports, hospitals, office buildings, and police stations.

“Usually when you develop an algorithm its nice to see where it succeeds, but you actually learn the most where you look at where the algorithm fails or it doesn't do well,” Pritt said. In trying to decipher why the algorithms struggled, Pritt said the competitors suggested that some objects simply don't have any distinguishing features from the point of view of a satellite image for the algorithms to recognize.

“Maybe the most important ingredient you need for these new types of algorithm to work is the dataset because these algorithms require a great amount of data to train on,” Pritt explained. “It's kind of analogous to the way a human will learn in childhood how to recognize things. You need lots of examples of what those things are and then you can start to generalize and make your own judgments,” he said.

But even with large amounts of training data that is correctly labeled, it is also possible the deep learning technology of today cannot reach the higher levels of intelligence to recognize nuanced differences. For example, Lockheed Martin's algorithm confused shipyards and ports 56 percent of the time. Pritt said that people “look at an image and they can tell that it's a port or a shipyard, they are usually looking at very subtle things such as if there is a ship in dry dock or if there is a certain type of crane present. They are looking for details in the image that are maybe higher level or more complicated than what these deep learning algorithms can do right now.”

However, the fact that these algorithms cannot do everything should not dismiss the significant contribution they could provide to the defense and intelligence community.

Hakjae Kim, IARPA's program manager for the fMoW challenge, said the benefits of this technology could extend far beyond faster image processing. “I want to look at it more in the perspective that we can do things we weren't able to do before,” Kim said. “Because its technology that we are now able to do x, y and z, there are more applications you can create because with the human power it is just impossible to do before.”

Kim and Pritt stressed managing expectations for CNN-based artificial intelligence.

“This is a real technology that will work, but it also has limitations. I don't want to express this technology as a magic box that will just solve everything magically,” Kim said. “I don't want the users in the field to get disappointed by the initial delivery of this technology and say 'Oh, this is another technology that was oversold and this is not something we can use," he added.

Part of managing our expectations for AI requires recognizing that although intelligence is in the name, this technology does not think and reason like humans. “A lot of the time we think that because we use the term AI, we tend to think these algorithms are like us, they are intelligent like us,” Pritt said. “And in someways they seem to mimic our intelligence, but when they fail we realize ‘Oh, this algorithm doesn't really know anything, [it] doesn't have any common sense.'”

So how are IARPA and Lockheed Martin working to improve their algorithms? For IARPA, Kim's team is working on updating and maintaining their dataset to ensure algorithms have the most up to date information to train on, ultimately making the CNN-based algorithms easier to trust. “[S]ubtle changes in the area mess up the brains of the system and that system will give you a totally wrong answer,” Kim explained. “So we have planned to continuously look over the area and make sure the algorithm we are developing and reassessing for the government to test on and use to be robust enough for their application," he furthered.

Work is also underway at American universities. Kim described how a team of researchers at Boston University are using the fMoW dataset and tested algorithms to create heat maps that visualize what part of the image algorithms are using to classify objects. They've found that sometimes it is not the object itself, but clues surrounding the object that aid most in classification. For example a “windmill that actually shows a shadow gives a really good indicator of what that object is,” Kim said. “Shadows show a better view of the object. A shadow is casting the side view of the object over on the ground, so [BU's heat map algorithm] actually points out the shadow is really important and the key feature to make the object identified as a windmill.”

But don't expect these algorithms to take away the jobs of analysts any time soon. “I think you still need a human doing the important judgments and kind of higher level thinking,” Pritt said. “I don't think AI will take away our jobs and replace humans, but I think what we have to do is figure out how to use them as a tool and how to use them efficiently, and that of course requires understanding what they do well and what they do poorly," he concluded.

https://www.c4isrnet.com/intel-geoint/2018/09/18/differentiating-a-port-from-a-shipyard-is-a-new-kind-of-problem-for-ai

On the same subject

December 27, 2018 | International, Naval

The US Navy’s surface fleet: Here’s what’s ahead in 2019
By: David B. Larter WASHINGTON — The U.S. surface fleet has a big year in store for 2019, and we're going to start getting more details very soon on what the future has in store for surface warriors. But surface leadership has been dropping clues on where things are going. Here's a handy reference guide for heading into January's Surface Navy Association annual symposium and for what the fleet has up its sleeve for the coming months. Robot wars The Chief of Naval Operations' Surface Navy Director Rear Adm. Ron Boxall forecast what was on his mind at a recent training and simulation conference in Orlando. The focus for Boxall and the N96 shop will be to get more sensors and weapons into the battlespace, distributed and networked, so that a smaller number of larger warships can act as command and control for smaller units. “If you think about what we are trying to do with the surface force, we have large and small surface combatants that will [ultimately make up part of the 355-ship Navy] but we have no requirement for unmanned surface vessels right now, which I see as an absolutely critical part of distributed lethality, distributed maritime operations environment that we are moving into,” Boxall said. “Ultimately I need more nodes out there.” N96 is looking closely at what might be needed for a large unmanned surface vessel, much like the Sea Hunter drone developed by the Defense Advanced Research Projects Agency. “I think these are what you need to go in the water and carry large things and be more places at less cost,” Boxall said. “So, in that nodal structure, we are looking at them becoming large sensors or large shooters, but we are still working out the requirement.” Developing unmanned vessels for military use was a key component of a recent agreement with NATO forged during the July summit. Future frigate Next year is the crucial year for FFG(X), the year when the Navy finalizes its requirements and puts the first hull out to bid for a 2020 detailed design and construction award. Look for news to creep out on this ship throughout the year but Boxall had some remarks on how it will fit into the fleet now in development. Boxall hinted that the planned 20 hulls may be a low-ball figure, and that he's looking to maybe keep the program going beyond that. “It will be a very capable ship, but it won't have a lot of capacity,” he said. “But it will be able to both sense and shoot and do command and control at a smaller level. It will be a much less expensive platform and I can have more of them.” Training Training is a major focus of Surface Navy boss Vice Adm. Richard Brown, and some ongoing efforts will start to bear fruit in 2019. Earlier in 2018, the Navy reprogrammed $24 million to build a Maritime Skills Training Program, which will be heavily reliant on simulators to bring together officer and enlisted watchstanders from both the bridge and the combat information center to train on equipment and work as a unit. “We've secured the funding for the maritime skills training centers, which is going to do two things: individual officer training through the [officer of the deck training],” Brown said earlier this year. “So that, in conjunction with building out the navigation, seamanship and ship-handling trainers in the fleet concentration areas, will give us integrated bridge and [combat information center] training at the high end. That's my No. 1 priority.” Those facilities will be ready for use by the waterfront in the 2021 time frame, Brown said. Upgrades to existing simulators are being rolled out this month. DDG-1000 Look for news on the future of DDG-1000. The first of the class, the Zumwalt, is wrapping up its combat systems installation in San Diego and will start the process of integrating the three-ship class into the fleet. We're going to find out more about its new mission – surface strike – and how the Navy plans to employ its behemoth new surface combatant. The Navy has pivoted away from its long odyssey to find a use for its advanced gun system, with requirements boss Vice Adm. William Merz saying in testimony in April that the Sisyphean task of getting a working gun on Zumwalt was holding the ship back. “Even at the high cost, we still weren't really getting what we had asked for,” he said. “So what we've elected to do is to separate the gun effort from the ship effort because we really got to the point where now we're holding up the ship.” Instead, efforts are going to focus on getting Zumwalt into the fleet and on the hunt for ships to kill. Large surface combatant Last up, the Large Surface Combatant should start getting some meat on its bones in 2019. Boxall and company are aiming to put the fleet on a course to buy its cruiser and destroyer replacement in 2023 or 2024, which means a request for information from industry could be in the near future. What we know is that, like the small surface combatant, the Navy wants commonality with other nodes in the network. That means a similar radar as on FFG(X), the same combat system and as much overlapping equipment as the fleet can manage to tamp down on compatibility issues and on how much specialized training sailors need to be on one platform or another. https://www.defensenews.com/naval/2018/12/26/the-us-navys-surface-fleet-heres-whats-ahead-in-2019/
April 8, 2022 | International, C4ISR, Security

Cyber defence agency gets significant boost in Liberals’ Budget 2022
Canada’s cyber defence agency gets almost $700 million over five years to bolster cyber defences in government and the private sector – and launch offensive ‘cyber operations.’
July 24, 2020 | International, Aerospace

NATO Airlift Management Program Expands C-17 Training Capability with Boeing Services Contract
Long Beach, California, July 23, 2020 - Boeing [NYSA: BA] will provide aircrew training support for the Strategic Airlift Capability's multinational fleet of C-17 aircraft under a new U.S. Air Force Foreign Military Sale contract for a C-17 Weapon System Trainer (WST) that will be stationed at Pápa Air Base, Hungary. “The NATO Airlift Management Programme provides Strategic Airlift Capability nations with full life cycle management for aircraft and facilities, embodying of collective ownership and Smart Defence,” said Gregory Clark, programme manager, NATO Airlift Management Programme. “The aircrew training partnership with Boeing will ensure ongoing readiness of C-17 aircrews to our member Nations.” Through partnership with Boeing, NATO and the Strategic Airlift Capability can rely on nearly 30 years of experience providing quality training systems and support for the global C-17 fleet. This is the first C-17 training device for Pápa Air Base. It will allow for agile aircrew and engine maintenance training and reduce the time and costs associated with travel, while ensuring aircrew readiness. “We are excited to continue our partnership with NATO and expand its C-17 aircrew training capability,” said Mark Hayes, Pápa, Hungary base manager for Boeing Global Services. “This regional training location will enable the Strategic Airlift Capability to add velocity to aircrew training and proficiency, maintain training readiness, and provide the flexibility to deliver European airlift for decades to come.” Boeing's C-17 WST includes an air vehicle station and loadmaster station, both with instructor operator stations; a core integrated processor task trainer; and computer-based training and courseware. Two years of initial spares to support the training device are also included. This C-17 training device includes innovative technology with upgrades to the visual system and glass mirror display system expanding the total field of view. Boeing is also incorporating technology enhancements to address obsolescence and concurrency with latest aircraft block configuration. NATO C-17 training at Pápa Air Base is expected to begin in 2022. About NSPA The NATO Support and Procurement Agency (NSPA) brings together, in a single organization, acquisition, logistic, medical and infrastructural capabilities including the related procurement function. As NATO's primary enabler, the Agency's mission is to provide effective and cost efficient multinational solutions to the Alliance, its thirty Nations and Partners. The NATO Airlift Management Programme (NAMP) Office located in Pápa (Hungary) is an integral part of the NATO Support and Procurement Agency (NSPA). It manages the full life cycle management of the C-17 aircraft stationed at Pápa Airbase, including acquisition and sustainment on behalf of the member Nations of the Strategic Airlift Capability (SAC). NSPA is headquartered in the Grand Duchy of Luxembourg, with main operational centres in France, Hungary and Italy. The Agency employs around 1,500 staff and overseeing over 2,500 contractors in NATO's missions across the world. More info: www.nspa.nato.int About Boeing Boeing is the world's largest aerospace company and leading provider of commercial airplanes, defense, space and security systems, and global services. As a top U.S. exporter, the company supports commercial and government customers in more than 150 countries. Building on a legacy of aerospace leadership, Boeing continues to lead in technology and innovation, deliver for its customers and invest in its people and future growth. # # # Contact Ekaterina Perdaris Boeing Communications +1 703-465-3532 +1 314-810-9592 ekaterina.g.perdaris@boeing.com Roger Gilles Boeing, EU & NATO Communications +32 476 860 954 roger.gilles@boeing.com Francesca De Florio NSPA Communications +352 3063 6309 +352 661 864 300 francesca.deflorio@nspa.nato.int View source version on Boeing Newsroom: https://boeing.mediaroom.com/NATO-Airlift-Management-Program-Expands-C-17-Training-Capability-with-Boeing-Services-Contract

All news

Differentiating a port from a shipyard is a new kind of problem for AI

On the same subject

Cyber defence agency gets significant boost in Liberals’ Budget 2022 Canada’s cyber defence agency gets almost $700 million over five years to bolster cyber defences in government and the private sector – and launch offensive ‘cyber operations.’

Cyber defence agency gets significant boost in Liberals’ Budget 2022
Canada’s cyber defence agency gets almost $700 million over five years to bolster cyber defences in government and the private sector – and launch offensive ‘cyber operations.’