The Intelligent Systems Laboratory (ISL) conducts groundbreaking research in three thrust areas: Complex Networks, Cyber-Physical Resilience and Brain-Machine Intelligence. Through research collaborations with our LLC Member companies, government (e.g., DARPA, IARPA, Dept. of Homeland Security, Office of Naval Research, Air Force Research Lab), commercial and leading academic institutions (25+), we create new and innovative capabilities for diverse applications such as cyber-security, unmanned autonomous systems, human performance augmentation, big data analytics, intelligence, surveillance, and reconnaissance, and electronic warfare.
ISL is advancing technology in three key areas:
We live in an interconnected world with smart devices, services, and systems generating massive amounts of heterogeneous and multi-modal data. Research in the Complex Networks thrust focuses on how to model and analyze an interconnected network of networks (NoN) framework so that we can extract useful information from the big data that these systems generate, as well as predict and influence their behavior.
The goal of our research in Complex Networks is to extract useful and actionable information from the deluge of multi-modal data generated by these interconnected devices, platforms and systems. In an NoN framework, we first develop techniques and models to discover and predict underlying dynamic relationships and structure of network entities (smart devices, platforms, systems, and people). We then develop algorithms to analyze and exploit the complex network structure and relationships, as well as methods to influence the network towards desirable outcomes. We are applying our research to cybersecurity, vehicle/platform health management, intelligence, surveillance, and reconnaissance (ISR), electronic warfare and social behavior analysis.
Interconnected systems, computers, vehicles and devices are all vulnerable to cyberattacks – with new attacks and vulnerabilities announced almost daily. Research in the Cyber-Physical Resilience thrust focuses on the automatic synthesis and verification of provably secure cryptographic protocols for data storage and computation, and on tools that automatically verify that system software adheres to specified security policies.
The goal of our research in Cyber-Physical Resilience is to protect computing devices, platforms (e.g., vehicles and planes), and critical infrastructure from cyberattacks, and to ensure the privacy and security of data computation and storage. We focus our research along two main directions: secure computation and automated verification and synthesis of secure software and cryptographic protocols. We are applying our technology to proactively secure cloud storage and control, secure and preserve privacy in databases and search of streaming data, secure distributed financial transactions, secure information flow in dynamic reactive systems, and build high-assurance automotive software.
The human brain is an amazing information processing device! Research in the Brain-Machine Intelligence thrust focuses on computational and mathematical modeling of cognition and brain functions to understand how humans process information and interact with the environment. This research will enable better training systems, advanced human decision aids, smarter autonomous systems and brain-like, low size, weight, and power processing systems.
The goal of our research in Brain-Machine Intelligence is to understand how humans process information and effectively interact with the environment to develop smarter autonomous systems, enhance human performance, and develop novel processing systems. Our research starts with the development of large-scale, neurobiologically and behaviorally faithful brain region models. We apply these models along with real-time measurements of brain activity (EEG, fMRI, fNIRS) to create advanced brain-computer interfaces, human decision aids, and neurostimulation-based enhanced training systems. We also use our models to develop novel, brain-inspired sensor exploitation, machine learning, and control algorithms, and to develop low size, weight, and power brain-based processing hardware for resilient autonomous systems, dexterous robots, and threat warning applications.
ISL develops a diverse set of technologies organized into four centers: Computational Network Intelligence, Secure and Resilient Systems, Human-Machine Cognition, and Autonomy Computing. The lab develops and demonstrates capabilities in autonomous driving, cyberphysical systems, and on-time prognostic tools and prototypes using data from open sources and our LLC members. ISL is also exploring various bio-inspired, non-Boolean processing methods including spike processing and analog pattern matching. Such novel computation paradigms will one day change the way computers process information, learn and yield more human-like decisions.
ISL develops novel technologies and customized solutions. Our goal is to discover the principles and mechanisms in the complex, emerging dynamics of interconnected human, machine, physical, and social ecosystems. We extract useful and actionable information, predict future events, and intervene and control cyberphysical networks with applications in vehicle/platform health management; intelligence, surveillance, and reconnaissance (ISR); cyber information warfare; and social behavior analysis networks. These applications sort and analyze large data from the deluge of multimodal, heterogeneous and streaming data generated by interconnected smart devices, platforms, and systems.
ISL's research goal in this area is ensuring security and reliability of computing resources that compose the internet of things (IoT). Combining techniques and theory from formal verification, program synthesis, cryptography, and distributed systems, we are developing tools for the synthesis and verification of high-assurance software. We also develop techniques for ensuring security and privacy of computations and data, and secure resilient protocols to provide reliable infrastructure in unpredictable adversarial environments.
Here ISL explores the brain's capability to learn, recall, adapt to uncertainty, and decide. Our goal is to change the relationship between humans and machines in two complementary directions. The first is to augment human performance by creating human-computer interfaces that sense cognitive and somatic states and adaptively apply neurostimulation. The second is to enhance machine intelligence based on neural learning and decision making to create cognitive processing systems. These thrusts enable applications in human-machine teaming, decision aids, threat detection, closed-loop training systems, adaptive autonomous systems, and dexterous robots.
The ISL goal is development of novel algorithms and software for specialized hardware, particularly applications requiring low size, weight, and power, and real-time processing. ISL also exploits novel computing paradigms implemented in emerging new devices, including neuromorphic chips. Target applications are autonomous driving, unmanned aerial and underwater systems, autonomous swarms, ISR, warfighter aiding, and IoT.
||||Future Consideration – ISL|
||||Causal Inference and Reasoning Scientist|
||||Machine Learning Engineer|
||||Machine Learning Scientist|
||||Machine Learning and Computer Vision Researcher|
||||Robotics and Machine Learning Engineer|
||||Formal Verification Scientist|
||||Future Consideration Internship|
||||Masters Intern in Robotics and Autonomy|
||||Doctorate Internship in distributed and multi-domain autonomous systems|
||||Summer Internship in Active Prediction|