AI accelerator chip
Artificial intelligence (AI) accelerator chips are semiconductor chips specifically designed to accelerate AI/ML computations and reduce energy consumption. These chips enable AI functionality on edge devices. The article argues that the speed of AI development may depend on the materials and design of new chips. In the short term, edge applications (such as high-fidelity computer vision on drone systems or real-time translation using natural language in head-mounted displays) may undergo significant changes. In the future, as chips become increasingly "brain-like," highly complex models, such as large language models, will be able to run on miniaturized, lightweight, and low-power devices.
Alternative technologies for positioning, navigation and timing
Positioning, navigation, and timing (PNT) alternatives offer a safer and more robust alternative to the existing Global Positioning System (GPS). This suite of technologies includes inertial, visual, low Earth orbit satellite, terrestrial radio frequency (TRF), and environmental and geophysical navigation and timing technologies. Future U.S. navigation systems are expected to incorporate multiple alternative PNT technology modalities and may integrate them using multimodal AI technologies.
Autonomous swarm robots
Autonomous swarm robots refer to a group of robots that work together as a whole to perform tasks. Applications of autonomous swarms include intelligence gathering, infrastructure monitoring, perimeter security maintenance, and target acquisition or kinetic strikes on the battlefield. Units within an autonomous swarm use AI "drivers" to communicate with humans, other units within the swarm, and non-swarm systems to optimize decision-making and execute tasks. The article points out that applications of autonomous swarms include inter-swarm combat, kinetic strikes against military personnel, equipment, and vehicles, and enhanced capabilities in intelligence, surveillance, reconnaissance, communications, electronic warfare, and logistical support.
Generative AI New Software
Generative AI software is a new type of AI/ML software used for reviewing, editing, and writing code, designed to reduce the burden of human coding and decrease vulnerabilities and security flaws. The article points out that the U.S. Department of Defense and other federal agencies are actively exploring ways to integrate generative coding technology. With advancements in generative coding tools, this technology will be able to handle more of the actual coding work, allowing humans to focus more on the design and logic of software rather than its programming rules.
High-density energy storage
High-density energy storage technology offers higher energy-to-weight ratio and energy-to-volume ratio compared to current energy systems. This technology can provide more efficient and longer-lasting power for a wide range of applications, from portable electronic devices to electric vehicles. Recent advances in silicon anode technology indicate that silicon anodes will offer better energy density and efficiency than graphite-based lithium-ion battery technology within the next 1-3 years. High-density energy storage is crucial for various missions, including supply chain logistics, base operations, and supporting frontline personnel. Future power systems will utilize pluggable silicon anode lithium-ion batteries and solid-state batteries, while hydrogen batteries hold long-term potential, but widespread adoption and full utilization are expected to take 10-20 years.
Hypersonic technology
Hypersonic technology enables aircraft or weapon systems to reach speeds of Mach 5 or higher. The U.S. military is developing two types of hypersonic weapons: cruise missiles equipped with air-breathing jet engines and high-speed gliders. Advances in digital technology have reduced the cost of hypersonic development and accelerated the pathways for innovation and expansion.
Multimodal AI
Multimodal AI is an AI system capable of simultaneously processing multiple data types, including text, images, audio, and video, aiming to better interpret context and make more accurate predictions. Multimodal fusion technology is increasingly being used to train AI models across multiple data types in fields such as language understanding, image generation, and biometrics. The article points out that the U.S. Department of Defense requires a large number of multimodal models to fuse different types of intelligence for decision-making output.
Non-kinetic anti-drone system
Non-kinetic anti-drone systems use highly concentrated energy (such as lasers, microwaves, and particle beams) to disrupt or destroy fast-moving threat targets. With the increasing prevalence of drones, future anti-drone systems will be equipped with AI-driven command and control systems capable of fusing sensor data, rapidly identifying new threats, and automatically making optimal responses under human supervision.
Post-quantum cryptography algorithms
Post-quantum cryptography algorithms refer to mathematical algorithms used to encrypt private communications of individuals, organizations, and governments. These algorithms are crucial for resisting attacks from current and future quantum computers. Post-quantum cryptography algorithms are based on complex mathematical models, and no quantum shortcuts have yet been discovered. The National Institute of Standards and Technology (NIST) is standardizing these models and advocating for a framework for cryptographic agility. The article argues that since the transition to post-quantum cryptography will take at least 10 years, relevant agencies urgently need to take immediate action. Cryptographic agility and hybrid cryptographic protocols will be key to adapting to emerging vulnerabilities.
Space situational awareness technology
Space situational awareness technology is used to gain a comprehensive understanding of the space operating environment, detecting threats and mitigating collision risks by studying and monitoring satellites and orbiting objects. This technology is achieved through a broad technology stack, including sensor modalities, computing resources, AI/ML and analytics, visualization technologies, and applications and services built upon these technologies. Future innovations will include the ability to accurately track and deploy spacecraft in higher orbits, the increase in space-based commercial sensors, and the adoption of augmented reality and virtual reality technologies.
You must be logged in to post a comment.