Centre for Applied Artificial Intelligence

As AI systems are deployed in high-stakes environments, such as health care, software development, telecommunications, manufacturing and public and private sector decision-making, we need to ensure that these AI systems remain trustworthy. Researchers at the University of Ottawa’s Faculty of Engineering are tackling this challenge by bridging the gap between foundational machine learning and real-world engineering applications, developing key tools for industry in the process.  

Research spans the full AI lifecycle—from data curation and model training to evaluation and deployment—and addresses key challenges, such as transparency, robustness and alignment. This helps answers the question: how trustworthy is this system, and for whom?  

Core areas include:  

• Data quality, bias, privacy and governance  
• Model interpretability and analysis of learned representations  
• Detection of failure modes, memorization and robustness gaps  
• Evaluation frameworks for fairness, safety and alignment

Researchers

Intelligent energy systems (IES) use AI and advanced information and communication technologies (ICT) to integrate renewable energy sources, storage technologies and smart grid infrastructure.  

These systems rely on continuous, bidirectional exchanges of energy and data across interconnected electricity, gas and thermal networks. AI-driven predictive models and adaptive control strategies enable dynamic balancing of supply and demand.  

IES help industrial partners improve operational efficiency, reduce costs and enhance energy resilience while supporting decarbonization objectives.  

Key capabilities include:  

• Real-time optimization of energy generation, distribution and consumption  
• Integration of variable renewable energy sources at scale  
• Demand-side management and peak load shifting  
• Cyber-secure, resilient grid operations 

Researchers

Artificial intelligence is transforming how decisions are made across critical fields, making ethical considerations central to system design, deployment and governance.  

Ethical principles—including fairness, transparency, accountability, safety and privacy—are integrated across both basic and applied AI. Emphasis is placed on real-world deployment, where AI systems directly affect individuals and society in areas such as health care, transportation, public infrastructure and national security.  

Key areas of focus include:  

• Responsible data collection, use and governance  
• Detection and mitigation of bias and sociotechnical risks  
• Evaluation of trade-offs between performance, safety and fairness  
• Human-centred design and stakeholder engagement  

This work is closely aligned with emerging regulatory frameworks and standards, including the EU AI Act and guidance from organizations such as NIST and ISO. It enables development of AI systems that are not only high performing, but also compliant, trustworthy and socially responsible.

Researchers

The Faculty of Engineering at the University of Ottawa is a recognized leader in applied AI for autonomous systems, robotics and intelligent cyber-physical infrastructures. Its interdisciplinary expertise spans AI, communications, control systems, sensing and cybersecurity to advance next-generation connected and autonomous technologies.  

Research focuses on systems operating in complex, dynamic environments, including:  

• Intelligent transportation and connected autonomous vehicles  
• Uncrewed aerial and ground systems  
• Swarm coordination and distributed autonomy  
• AI-enabled decision-making and situational awareness  

This work is supported by advanced infrastructure, including autonomous system testbeds, UAV platforms, simulation environments, and robotics and control laboratories, enabling both development and real-world validation.  

The Faculty contributes to major collaborative initiatives at the national and provincial levels and plays an active role in developing international standards (e.g., ISO, UL, SOTIF, 3GPP) to ensure alignment with safety and interoperability requirements.

Researchers

Integrating AI with sensors, embedded processing and communication technologies, these systems analyze complex signals, detect patterns and predict future states from continuous measurement streams. These sensors extend beyond data acquisition to enable context-aware interpretation, learning and decision-making. They provide accurate measurements along with uncertainty quantification and traceable calibration, supporting reliable operation in highly automated environments.  

Key capabilities include:  

• Fusion of traditional sensors with embedded intelligence and virtual (soft) sensing models  
• Real-time signal analysis, predictive modelling and anomaly detection  
• End-to-end integration from measurement to decision-making  

Applications span a wide range of fields:  

• Industrial systems: real-time process analytics, inline quality assurance, predictive maintenance and fault detection  
• Infrastructure: structural health monitoring of bridges, buildings and critical facilities  
• Environment and smart cities: real-time monitoring of air quality, water systems and urban infrastructure  
• Health care: continuous physiological monitoring using wearable sensors for early detection and remote care

Researchers

As AI becomes increasingly embedded in our critical infrastructure, telecommunications, health care and national security, cybersecurity and privacy present increasingly complex and urgent challenges. At the Faculty of Engineering, a strong cluster of expertise addresses these challenges across the full spectrum of concern, from theoretical foundations to operational deployment.  

At the heart of this work is the uOttawa–IBM Cyber Range, a dedicated facility for simulating real-world cyberattack scenarios, which allows researchers, students and industry partners to train and test defences under realistic conditions.  

Research areas include:  

• AI-driven detection of phishing, malware and intrusions  
• Vulnerability detection and anomaly analysis in large-scale systems  
• AI-based methods to secure IoT and embedded systems  
• Protection of machine learning systems against adversarial attacks, data poisoning and model manipulation  
• Theoretical work on cryptographic foundations — including post-quantum cryptography, pseudo randomness and privacy-enhancing technologies

Researchers

Making AI-powered software systems that practitioners can trust is one of the field’s most pressing challenges. These systems require the same rigour as traditional engineering systems to ensure reliability and trustworthiness.  

This work combines modern AI techniques—including large language models (LLMs), natural language processing (NLP) and machine learning— with established software engineering practices, such as formal specifications, modeling and verification.  

This integrated approach supports the development of high-performance software systems that meet both technical and regulatory requirements.  

Technical focus areas include:  

• AI-assisted requirements engineering and system design  
• Reliable, AI-assisted code generation  
• User experience design using AI and user experience design for AI-enhanced interfaces  
• Automated testing and verification of complex and autonomous systems  
• Safety assurance for AI-enabled software

Researchers

Natural language processing (NLP) enables advanced analysis and understanding of unstructured text and multimodal data.  

Expertise spans both statistical and deep learning approaches, with a focus on:  

• Large language model (LLM) fine-tuning and evaluation  
• Text classification, information retrieval and lexical semantics  
• Bias detection, mitigation and explainability
• Retrieval-augmented generation and safety evaluation  

This work is applied in high-stakes areas such as health care and law, including collaborations with hospitals, research institutes and industry partners.  

Key applications include:  

• Detection of harmful or high-risk language (e.g., hate speech, mental health indicators)  
• Automated screening and analysis of clinical and patient-generated text  
• Identification of risks in LLM deployment and fine-tuning
• Explainable emotion and sentiment classification

In addition, NLP methods are applied to engineering and scientific areas, supporting tasks such as requirements analysis, automated test generation and regulatory compliance.

Researchers

Wireless networks can be considered the information highways of modern societies: they provide connectivity for users, robots and AI agents. Every wireless network generation, from 2G to 5G, has increased capacity, serving more bits to more users.  

Research at the intersection of AI and wireless systems focuses on applying advanced machine learning techniques to optimize systems, improve network control and judiciously allocate resources.  

These capabilities are critical in industries that deploy connected infrastructure, autonomous systems and mission-critical communications.  

Key areas include:  

• AI-driven network optimization under dynamic conditions  
• Autonomous network management and orchestration  
• Support for ultra-dense IoT deployments and edge systems  
• Integration of AI into communication protocols and control layers

Researchers

• Chen, Nan
• Kantarci, Burak
• Yao, Jianping

Research on critical infrastructure focuses on improving the durability, resilience and lifecycle performance of engineering systems through integrated approaches combining materials science, structural engineering and data-driven methods.  

This work addresses key challenges related to material degradation, environmental exposure and long-term system performance, while advancing sustainable, low-carbon solutions.  

Artificial intelligence plays an increasing role in:  

• Interpreting complex inspection and monitoring data  
• Supporting predictive modelling of degradation and failure  
• Automating advanced characterization methods, including visual and microscopic analysis  

These capabilities improve the speed, consistency and objectivity of infrastructure assessment and asset management.  

By linking experimental data, field performance and digital technologies, this research enables more reliable, efficient and sustainable management of critical infrastructure systems across industry and the public sector.

Researchers

Novel approaches that leverage AI are reshaping the discovery and production of materials across industrial sectors, driven by the need to lower costs and improve scalability. AI is being applied to material production in fields ranging from energy and aerospace to automotive and defence, with a focus on sustainability, hydrogen production and hydrogen storage.  

This work integrates AI simulation, density functional theory (DFT) validation, experimental synthesis, and state-of-the-art characterization within a closed-loop optimization framework.  

This holistic approach enables rapid iteration between modeling and experimentation, which significantly reduces research and development cycles for complex materials, components and systems.  

Key capabilities include:  

• Forward AI design for predicting properties of known systems using large-scale data and high-performance computing  
• Inverse AI design for discovering novel materials with targeted properties and performance  
• Optimization of chemical composition, crystal structure, microstructure and synthesis parameters  
• Application of advanced methods, including deep learning, evolutionary algorithms, metric learning and adaptive learning  

Researchers

Our current research activities in AI for health care span multiple areas, including human movement analysis, biomedical evidence processing, agentic health-care systems, affective computing and remote physiological sensing.  

Key capabilities include:  

• Human movement analysis: markerless motion capture from video, statistical shape modelling for joint health assessment and predictive simulation of gait using evolutionary algorithms  
• Biomedical evidence processing: automated search, extraction and synthesis of scientific and clinical literature, producing structured, actionable insights and reducing  research waste  
• Agentic health-care systems: AI-driven conversational and recommender tools supporting triage, care navigation and clinical decision-making with human oversight  
• Multimodal health monitoring: analysis of language, behaviour and wearable sensor data for affect recognition, physiological sensing and mental health assessment

Researchers

Biopharmaceuticals, ranging from insulin to next-generation cancer therapies, are produced using living cells. While they are transforming modern health care, manufacturing them remains complex, resource-intensive and costly. At the University of Ottawa, we are advancing machine learning approaches to improve the efficiency, scalability and reliability of biomanufacturing processes.  

Key capabilities include:  

• Data-driven simulation: Modelling complex bioprocess dynamics using time-series machine learning to predict full process trajectories, optimize operating conditions and improve production efficiency  
• Scale-up modelling: Applying deep learning to historical data to accurately predict process behaviour at industrial scale, reducing reliance on costly trial-and-error experimentation  
• Digital twins: Converting real-time bioreactor sensor data into virtual models that enable monitoring, forecasting and automated control of key variables such as nutrient levels, cell density and product concentration

Researchers

• Chahal, Shawn
• Wang, Erika