Shades of grey: Open-source data blur the lines between geographic, ideological, and cyber targets
Part of modern warfare is leveraging the abundance of information being gathered in the digital age to actionable intelligence. Anika Torruella investigates the challenges of composing complex analysis in 'grey' environments based on enormous amounts of open-source, non-classified, and other raw intelligence data.
Open-source technologies have an undeniable role to play in transforming communication and defence industry processes. Cloud computing, wireless sensor networks, data-driven automation, machine-to-machine communication, and smart or adaptive systems promise to do more than improve operational efficiency.
"Obviously the threat landscape is changing," Keith Johnson, chief technology officer of Leidos Defense and Intelligence Group told Jane's. "Threats are not necessarily as stationary as they once were. They're evolving. The tradecraft of our adversaries is changing, and so we need to be more agile to speak to these new threats."
"We are going to have to rethink how we do security, how we do this integration, and how do we protect from digital errors in the system. Ultimately, we are all moving to a greyer environment,"he added.
Above - Members of the US Naval Postgraduate School's Common Operational Research Environment (CORE) Lab in 2012 who worked on innovative ways to analyse open-source information gleaned from social network sites. (US Navy/Kenneth Stewart)
"The potential of artificial intelligence [AI] and machine learning [ML] for the [US] Department of Defense [DoD] is particularly evident in three areas: intelligence, surveillance, and reconnaissance [ISR]; cybersecurity; and healthcare," Major Jamie Davis, spokesperson for the US Office of the Assistant Secretary of Defense told Jane's.
"ML and AI are key enablers, and a top request from field commanders is more ISR. Their objective is to maximise lethality and capability with responsive ISR capabilities supporting near-realtime decision-making. You stand back and you look at Amazon, who really in my view started user analytics, and what they do with data [to drive recommendation algorithms], ... I think what defence has done and military operations in general [are doing] is taking that capability and transferring it into their applications."
Jane Chappell, vice-president of the Global Intelligence Solutions (GIS) mission area of Raytheon Intelligence, Information, and Services (IIS) agreed. "Quite frankly, there is a lot of that work going on. There is a lot of focus on making sure that defence takes advantage of what is happening in commercial applications and the innovation that is happening there, and making sure that that gets applied towards their mission."
Although large internet companies, such as Facebook or LinkedIn, have been the pioneers of leveraging open-source data to drive algorithms and software architecture, militaries are looking into open-source data that incorporate many more sources than social media intelligence (SOCMINT).
Another challenge is the deficit of personnel that have been trained in managing systems that process enormous amounts of data at a high speed, and the closed-off nature of military analytics tools can make aggregating data across departments difficult. In addition, militaries want to ensure that critical decisions stay in the hands of human decision makers as the machine elements in this complex chain of systems grow and are exponentially faster than the human elements.
"Sometimes, we are not doing enough because we don't have enough people, because the processes today are fairly manual," Johnson said. "One way we can do more is to insert more automated analysis into that cycle."
"It is going to be the integration of machine-to-machine communications, artificial intelligence, and a better machine-to-human connection that's going to help us maintain our edge," Johnson added. "But it is also going to be dependent on our ability to 'break the stovepipes' of the systems that we use to enable better integration. Better integration of data across those systems and then also integration of new data sources, like the open-source data sources."
Above - 3D cloud rendering of the Bird's Nest stadium in Beijing, China, created by Raytheon's Intersect Dimension software, which aggregates passive 2D and other remote sensing imagery sources to create multi-dimensional models. (Raytheon)
Some of the products that Raytheon has developed in this area are in its Intersect family. For instance, Intersect Dimension uses open-source 2D and remote sensing imagery to rapidly create multi-dimensional models for projects such as disaster relief, immersive training, and road monitoring. Intersect Reveal automates basic full-motion video analysis tasks, such as registration, tracking, classification, and indexing; and combines the results with multi-intelligence (MULTINT) data in near-realtime. "But it is really taking in different types of data, across multiple domains, and making sure that you have the right context and that your data that you are using for decisions are in context of the overall environment," said Chappell.
Above - US Chief of Naval Operations (CNO) Admiral John Richardson tours the pilot house control centre of littoral combat ship (LCS) USS Independence (LCS 2) in 2016.
Lockheed Martin built sensors, cameras, and software systems on board LCSs to collect open-source data from propulsion, electrical, engineering control, and auxiliary systems to predict mechanical issues and determine causes. The data feeds from equipment are beamed hourly via satellites to US Navy and industry teams for analysis. (US Navy/Nathan Laird)
The IOT represents a particular challenge as a convergence of software, hardware, and the electromagnetic (EM) spectrum; and encompasses applications such as mobile computing, communications, software architectures, sensors, and data analytics.
"There is a blurring of the threat," Johnson said, regarding the new challenges presented by networked systems based on open-source data. "Everything is a network, and understanding the connections between people and capabilities and ideologies and intent is more important now than it used to be, when things were more static and we understood things. That requires more nuance and more alternative views, and that's really where machine intelligence can help in helping us put the pieces together in different ways."
Chappell also explained that cloud computing, in particular, was an important technology to deal with large volumes of data and enable faster sharing of asymmetric data types for decision makers."You know as you continue to have access to more and more data and need to process that data, it's important to have the elasticity that comes with the cloud," Chappell said.
"Cloud computing or virtualising ... can augment [tactical systems]," Johnson agreed. "Cloud is a term that definitely needs some definition because you can have a private cloud or even a tactical cloud where everything is closed and cyber-hardened and deployed in a military network."
"That is good because it allows those mission-critical applications and systems to more efficiently use the hardware resources. Cloud in general just allows you to leverage hardware resources more efficiently."
Jane's Consulting works with OSINT professionals around the world to optimise their OSINT processes and practices. Learn more about our training workshops here.
Jane's Editorial Staff
Posted 14 July 2017
- US Navy deploys support in search for downed Japanese F-35A pilot
- Extreme right activists build transnational narratives online
- Brazilian defence programmes face lowest acquisition budget in ten years
- The threat from far-right extremism: rising, less predictable and hard to determine
- Highest pay raise in ten years for military personnel, reveals US FY20 defense budget
- First terror attack recorded in New Zealand | Jane’s Terrorism and Insurgency Centre analysis
- The results are in…the winners of the Jane’s ATC Awards have been revealed!
- Packages sent to UK transport hubs claimed by IRA in bid to create unease