Research

Center for Artificial Intelligence and Research (CAIR) Laboratory

Advancing Physical AI, Interactive AI, and Intelligent Connected Systems

The Center for Artificial Intelligence and Research (CAIR) conducts interdisciplinary research in artificial intelligence systems that operate across physical, digital, and connected environments. Located in Room 213, Pangborn Hall, CAIR provides a collaborative research environment for doctoral scholars, faculty, and partners working at the intersection of AI, engineering, computing, and applied innovation.

CAIR’s research builds on core strengths in signal processing, computer vision, multimodal artificial intelligence, autonomous systems, sensor fusion, language modeling, and intelligent robotics. Our work focuses on developing AI systems that can interpret complex data, reason across multiple sources of information, interact with humans, and support intelligent action in real-world settings.

Rather than focusing only on software-based AI, CAIR emphasizes AI that is connected to the world around it - AI that can perceive through sensors, understand context, support decisions, and operate reliably in dynamic environments.

Physical AI

Physical AI refers to intelligent systems that are embedded in, connected to, or directly involved with the physical world. At CAIR, this includes robotics, autonomous platforms, sensing systems, cyber-physical environments, smart infrastructure, and AI-enabled monitoring systems.

This research direction focuses on building AI systems that can sense their surroundings, interpret changing conditions, and support intelligent action. These systems may operate in environments where data is incomplete, noisy, fast-changing, or distributed across multiple devices and sensors.

CAIR’s Physical AI research includes:

• Robotics and autonomous systems

• Vision-guided perception

• Edge autonomy and real-time decision-making

• AI-enabled cyber-physical systems

• Smart infrastructure monitoring

• Human-robot collaboration

• Simulation environments and digital twins

• Autonomous platforms for complex environments

Potential applications include intelligent transportation, smart cities, defense and security systems, infrastructure resilience, logistics, precision healthcare, manufacturing, and environmental monitoring.

Interactive AI

Interactive AI focuses on systems that work with people rather than simply producing outputs. CAIR develops AI technologies that can support human decision-making, improve learning, assist with complex tasks, and provide explanations that users can understand and trust.

This research area includes human-centered AI systems that can engage through language, vision, sound, and other forms of multimodal interaction. The goal is to create AI that is not only accurate, but also usable, transparent, adaptive, and aligned with human needs.

CAIR’s Interactive AI research includes:

• Human-AI teaming

• Explainable and trustworthy AI

• Intelligent dialogue systems

• Multimodal human-computer interaction

• Adaptive learning systems

• AI-supported education and training

• Decision-support systems

• Responsible and human-centered AI

This work is especially relevant for education, healthcare, workforce development, public safety, research support, and mission-critical decision-making environments where human oversight and trust are essential.

Intelligent Connected Systems and AIoT

CAIR studies intelligent connected systems that combine AI with networks of devices, sensors, edge platforms, and smart environments. This includes work related to the Artificial Intelligence of Things (AIoT), where intelligence is distributed across connected physical and digital systems.

In these environments, AI must often operate close to where data is generated. This requires systems that can process information efficiently, respond in real time, protect sensitive data, and coordinate across multiple devices or platforms.

CAIR’s research in intelligent connected systems includes:

• Edge AI and distributed intelligence

• Smart sensing networks

• Secure AI-enabled IoT systems

• Connected healthcare technologies

• Intelligent transportation systems

• Smart infrastructure and built environments

• Real-time monitoring, detection, and response

• Privacy-aware and secure AI deployment

This research supports future systems that are more responsive, resilient, efficient, and secure across healthcare, infrastructure, transportation, manufacturing, defense, and public services.

Real-Time Multimodal Perception

A major strength of CAIR is the development of AI systems that can interpret information from multiple sensing sources. Real-world environments are complex, and no single data type is always sufficient. By combining visual, acoustic, radar, textual, and sensor-based information, CAIR develops systems with richer situational awareness.

This research supports:

• Object detection and tracking

• Sensor fusion in dynamic environments

• Audio-visual understanding

• Radar and vision integration

• Multimodal situational awareness

• Detection and localization under uncertainty

• AI perception for robotics and autonomous platforms

These capabilities are important for systems that must operate in environments affected by noise, motion, limited visibility, uncertainty, or rapidly changing conditions.

Signal, Image, and Acoustic Processing

CAIR’s work in AI is strongly grounded in advanced signal and image processing. These foundations support applications that require accurate detection, classification, restoration, localization, and interpretation of complex signals.

Research areas include:

• Chemical agent detection and classification

• Radar signal processing and target tracking

• Restoration and enhancement of image sequences

• Acoustic modeling, detection, and localization

• Biomedical and environmental signal analysis

• Image processing for surveillance, healthcare, and infrastructure applications

This research provides the technical foundation for intelligent systems that must understand complex physical signals and convert them into reliable, actionable information.

Autonomous and Agentic AI Systems

CAIR also investigates AI systems that can plan, reason, communicate, and adapt over time. This includes private and domain-specific language models, agent-based AI systems, and architectures designed for autonomous workflows.

The goal is to create AI systems that can assist with complex tasks while remaining reliable, secure, and aligned with human goals.

Research areas include:

• Private and domain-specific large language models

• Agent-based AI architectures

• Autonomous planning and task execution

• Adaptive decision-making

• Intelligent dialogue management

• AI assistants for research and operational workflows

• Safe and responsible deployment of autonomous AI systems

These systems can support applications in research, education, engineering, healthcare, administration, and mission-oriented environments.

Retrieval-Augmented and Memory-Enhanced AI

CAIR develops AI systems that can use external knowledge, retrieve relevant information, and maintain context over time. This is especially important for specialized domains where accuracy, traceability, and contextual awareness are essential.

Research areas include:

• Retrieval-augmented generation

• Long-term memory architectures

• Knowledge-grounded reasoning

• Context-aware intelligent assistants

• Domain-specific AI systems

• Secure and private knowledge systems

These technologies help AI systems provide more reliable, informed, and useful responses in complex domains such as healthcare, education, engineering, research, and decision support.

AI-Driven Robotics

CAIR’s robotics research integrates perception, reasoning, language, control, and action. By combining computer vision, signal processing, multimodal AI, sensor fusion, and autonomous planning, CAIR develops robotic systems that can operate more effectively in real-world environments.

Research areas include:

• Vision-guided robotics

• Human-robot interaction

• Autonomous navigation

• Robotic perception and control

• Multimodal robotic intelligence

• Edge AI for robotic systems

This research supports the development of robotic systems that are more adaptive, interactive, and capable of functioning in complex and changing environments.

Research Impact

CAIR’s research is designed to support practical and mission-oriented applications across:

• Healthcare

• Education

• Smart infrastructure

• Transportation

• Manufacturing

• Robotics

• Defense and security

• Environmental monitoring

• Public safety

• Human-AI collaboration

Across all areas, CAIR is committed to developing AI systems that are not only technically advanced, but also trustworthy, responsible, secure, and deployable.