The Defense Advanced Research Projects Agency (DARPA) has always been at the forefront of innovation, pushing the boundaries of technology to enhance the capabilities of the United States military. In recent years, DARPA’s focus on developing artificial intelligence (AI) has led to groundbreaking advancements, particularly in the field of adaptive control for military systems. DARPA’s Learning Intelligent Control (LINC) program aims to revolutionize the way military systems operate by enabling them to adapt dynamically to changing missions and theaters. In this article, we will explore the significance of the DARPA LINC initiative and its potential impact on the future of military operations.
The Need for Adaptive Control
A control system is a system that manages, commands direct, or regulates the behavior of other devices or systems. It is used to achieve a desired output or performance by monitoring and adjusting inputs and outputs to meet predetermined specifications or objectives.
In general, a control system consists of several interconnected components, including sensors, controllers, actuators, and feedback mechanisms. Sensors are used to measure various inputs, such as temperature, pressure, or position. Controllers then use this information to make decisions about how to adjust the system’s behavior to achieve the desired output. Actuators are used to make the necessary adjustments to the system’s inputs and outputs, and feedback mechanisms are used to monitor the system’s behavior and adjust the control signals as needed.
Traditional control systems are based on mathematical models of the system and rely on a set of pre-defined rules or algorithms to regulate the system’s behavior. These rules are typically based on assumptions about the system’s behavior, which may not always hold true, and may require manual adjustments to maintain optimal performance.
At design time, these systems are built to handle a range of expected operating environments and parameters. Adapting them is currently done in an improvisational manner – often involving custom-tailored aftermarket remedies, which are not always commonly available, require a skilled technician to install, and can take months or even years to procure. Further, as they evolve and are placed outside of their original design envelop these systems can fail unexpectedly or become unintentionally dangerous.
Control design currently aims to model the range of operating environments that are anticipated at design time. Plans can fail, however, when physical attacks, unforeseen conditions, or unanticipated use places the system outside the design envelope.
Adaptive control systems, on the other hand, are designed to adjust and optimize their performance based on real-time feedback from the system. These systems use sensors and other feedback mechanisms to continuously monitor the system’s behavior and make adjustments accordingly. By constantly adapting to changing conditions, adaptive control systems can improve system performance and maintain optimal operation under varying conditions.
For deeper understanding of Adaptive control and applications please visit: Adaptive Control: Theory, Applications, Simulation, and Analysis
Adaptive Control in Military Systems
Many complex, cyber-physical military systems are designed to last for decades but their expected functionality and capabilities will likely evolve over time, prompting a need for modifications and adaptation. High Mobility Multipurpose Wheeled Vehicles (HMMWV), for example, had a design life of 15 years, but are now undergoing modernization to extend the fleet’s average age to 37+ years.
Traditionally, military systems have been designed for specific tasks and environments, making them less flexible when confronted with rapidly changing scenarios. In modern warfare, missions can shift swiftly, and theaters of operation can vary greatly, requiring the ability to adapt on the fly. This necessity has prompted DARPA to invest in developing AI-driven adaptive control systems that can optimize military assets’ performance across a wide range of circumstances.
The Department of Defense (DoD) systems are particularly long-lived, so ongoing adaptation would permit continual modification as missions and theaters change, providing a strategic advantage over an adversary. “Today, we start with exquisitely built control systems but then someone needs to add something or make a modification – all of which results in changes to the safe operating limits,” said DARPA program manager John-Francis Mergen. “These changes are done in a way that wasn’t anticipated – or more likely couldn’t have been anticipated – by the original designers. Knowing that military systems will undoubtedly need to be altered, we need greater adaptability.”
AI can be used in adaptive control to provide more accurate and effective control. One way AI can be used in adaptive control is through machine learning algorithms that can learn from data and adjust control parameters accordingly. For example, in a manufacturing process, sensors can be used to collect data on various aspects of the process, such as temperature, pressure, and flow rate. This data can then be used to train a machine learning algorithm to predict the behavior of the process, and the algorithm can be used to adjust the control parameters in real time to optimize performance.
Another way AI can be used in adaptive control is through reinforcement learning, where the control system is treated as an agent that interacts with its environment and learns from its actions. In this approach, the control system is rewarded for achieving desired outcomes, and over time it learns to adjust its actions to maximize the rewards.
The LINC Program: Unleashing AI’s Power
DARPA launched the Learning Introspective Control (LINC) program in August 2021. The program aims to develop machine learning (ML)-based introspection technologies that enable systems to adapt their control laws when encountering uncertainty or unexpected events.
The program also seeks to develop technologies to communicate these changes to a human or AI operator while retaining operator confidence and ensuring continuity of operations.
The DARPA LINC program focuses on integrating AI technologies into military systems to enhance their decision-making capabilities. By leveraging machine learning and advanced algorithms, LINC enables military assets to learn from experience, interpret complex data, and respond in real-time to unforeseen challenges.
A. Learning from Experience: LINC aims to create systems that can accumulate knowledge from previous missions and apply it to new situations. By learning from both successes and failures, military systems become more efficient and effective in tackling evolving threats.
B. Real-Time Decision-Making: In dynamic operational environments, split-second decisions can have profound consequences. LINC’s AI-driven adaptive control empowers military systems to analyze vast amounts of data and make informed decisions in real-time, reducing response times and increasing mission success rates.
C. Robustness and Flexibility: The adaptability of military systems is crucial for maintaining a tactical advantage. LINC’s AI-based adaptive control allows assets to adjust their strategies, tactics, and responses to suit various scenarios and emerging threats.
“When a system ‘wakes up’ in a different space, it needs to be able to realize there are things it can’t do anymore or new things it can, and ‘learn’ how to adapt to its new operating reality,” noted Mergen. “With LINC, we want to provide physical systems with the ability to figure out what is still feasible, alert the operator, and then help them operate in that new space.”
Developing LINC technologies will require addressing a specific set of challenges related to learning control and communicating situational awareness to the operator. Current state of the art (SOTA) ML approaches are not robust to unknown or unstructured parameter uncertainty, owing largely to the bounds set on their operation at design time as well as their reliance on fixed assumptions about their operating model. Further, complex systems – like drone swarms – are unable to rapidly converge on a common solution. When damage occurs to a single drone, the swarm is unable to uniformly adapt, potentially resulting in a failed operator or unsafe operating conditions.
Advantages of LINC
A. Improved Mission Success: With the ability to adapt to changing situations, military systems equipped with LINC can maintain their effectiveness across a wide array of missions and theaters, ensuring higher success rates.
B. Reduced Risk to Personnel: By utilizing AI-driven adaptive control, military operations can become more autonomous, reducing the exposure of human personnel to potentially hazardous situations.
C. Enhanced Efficiency: LINC enables military systems to optimize resource utilization, thus minimizing waste and ensuring that each asset is utilized to its fullest potential.
D. Scalability and Interoperability: DARPA’s LINC program aims to develop solutions that are scalable and interoperable across different military platforms, facilitating seamless integration into existing and future defense systems.
LINC will be a four-year, three-phase program; the first phase will last for 18 months, and the second and third phases will last for 156 months each.
Initial work will involve an iRobot PackBot and a remote 24-core processor. This ground robot weighs 20 pounds; measures 26.8 by 15.9 by 7.1 inches; has tracked and untracked flippers; moves at 4.5 miles per hour, and operates in temperatures from -20 to 50 degrees Celsius.
The remote processor has an Nvidia Jetson TX2 general-purpose graphics processing unit (GPGPU), dual-core NVIDIA Denver central processor, Quad-Core ARM Cortex-A57 MPCore processor; 256 CUDA software cores, eight gigabytes of 128-bit LPDDR4 memory, and 32 gigabytes of eMMC 5.1 data storage.
A key goal of the program is to establish an open-standards-based, multi-source, plug-and-play architecture that allows for interoperability and integration — including the ability to easily add, remove, substitute, and modify software and hardware components quickly.
LINC is seeking to achieve its goals in three main research areas.
- “LINC’s first research area will seek to overcome existing limitations in learning models and ML techniques that currently hamper system adaptation.” The program will explore how to provide a system with the ability to sense change and then reconstitute control using only onboard sensors and actuators. LINC aims to develop new control regimes that detect and characterize changes in the system’s operations in real-time, rapidly find solutions for reconstituting control under these changing conditions, and then calculate operating limits to identify a safe operating envelope.
“The idea is that you have a plethora of indigenous sensors on the system, and you can use these to determine and define a new set of control laws. With those new laws, you can then calibrate the system,” said Mergen.
- “A second research area will focus on improving how situational awareness and guidance are shared with the operator.” Another challenge area LINC seeks to address is around operator communications. Today, operators are not often provided with sufficient explanations or guidance around a system’s behavior or it’s situation-specific operating limits. Existing cues to operators about system dynamics don’t always provide options, making it difficult for an operator to appropriately trust the information its receiving. Further, interpreting current system diagnostics displays, which are not always intuitive, creates additional cognitive load for human operators. This further erodes operator trust and can lead to misunderstanding, confusion, and incorrect actions.
A second research area will focus on improving how situational awareness and guidance are shared with the operator. This area will explore ways of translating and effectively communicating the operational information generated by the dynamic model developed under the first research area. The resulting technologies must be able to provide the operator – whether human or AI – with updates on the operating status of the system as well as cues for safe actions. Further, they must be able to help retain operator trust by providing optionality and explainability around what’s happening “under the hood.”
- “A third research area will focus on testing and evaluating the resulting technologies.” LINC expects to use demonstration platforms that will evolve in sophistication and complexity throughout the life of the program – starting with a realistic physical model and progressing to a military-relevant system in the program’s final phase.
Peraton Labs to Develop AI-Powered Adaptive Control System for DARPA
The Defense Advanced Research Projects Agency has selected Peraton‘s applied research unit to create machine learning-based introspection tools to enable military assets to respond and adapt to unpredictable events.
Peraton Labs said in Feb 2023 it will develop an artificial intelligence-powered adaptive control system that will provide guidance and situational awareness in real-time to an autonomous controller or human operator under the Learning Introspective Control program.
Petros Mouchtaris, president of Peraton Labs, said the Adaptive Control with AI tool will help machine and human operators maintain system control during adverse situations.
He added that the real-time adaptive control technology will offer ground vehicles, ships, unmanned aerial vehicles, robotic systems and other physical assets with “substantial advantages in speed, agility, and responsiveness for operations in dynamic theater environments.”
Ethical and Security Considerations
While the prospects of AI-based adaptive control are promising, they also raise ethical and security concerns. DARPA acknowledges these concerns and is committed to ensuring that the technology is developed and deployed responsibly.
A. Human Oversight: DARPA emphasizes the importance of human operators maintaining control and oversight over military systems equipped with AI. Decisions with ethical implications should remain in human hands to prevent unintended consequences.
B. Cybersecurity: As with any AI system, robust cybersecurity measures must be in place to prevent potential adversaries from exploiting vulnerabilities in AI-driven military systems.
The DARPA LINC initiative represents a significant leap forward in the field of AI-based adaptive control for military systems. By creating intelligent and flexible assets that can adapt to changing missions and theaters, DARPA is providing the United States military with a powerful tool to maintain its competitive edge in an ever-evolving global security landscape. However, it is essential to approach this technology with caution and ensure that ethical guidelines are followed, safeguarding against any unintended consequences. The potential benefits of LINC are vast, and with responsible development and deployment, the program holds the key to reshaping the future of military operations.