RADS Research Day 2025

• In-person event

• Contact
• Nafiseh Kahani, kahani@sce.carleton.ca

Join us for our annual RADS Research Day! This year’s event will be in person in the Carleton University Engineering Design Centre (EDC) and the Systems and Computer Engineering Maker Lab (ME 4463). This year’s event features an exciting program consisting of student lightning talks and poster presentations, as well as an invited talk from Dr. Imran Ahmad.

Agenda

Invited Talk

Title: The Anatomy of Action: How AI Agents Reason, Plan, and Execute
Presenter: Dr. Imran Ahmad
Bio: Dr. Imran Ahmad is a Data Scientist specializing in big data, machine learning, and artificial intelligence. With over 20 years of experience in analytics and technology training, he has led many projects in predictive modeling and deep learning for the Canadian Federal Government and Cloudnum Inc. He holds a Ph.D. in Systems and Computer Engineering from Carleton University and is the bestselling author of 50 Algorithms Every Programmer Should Know.<\br>

Abstract: While we’ve seen rapid advancements in generative AI that responds, the next frontier is agentic AI that acts. This 40-minute presentation moves beyond the chatbot to explore the rise of autonomous AI agents. We will dissect the core components that allow an agent to perceive, reason, and interact with digital environments to achieve complex goals. This session is designed for graduate students and will provide a technical deep-dive into the “thinking” loop of modern agents, a live demo of an agent in action, and an overview of the powerful frameworks (like LangChain) used to build them. Attendees will leave with a clear understanding of the current agentic landscape and the key research challenges that are defining the future of AI.

Presentation Agenda (40 Minutes)

• The Agentic Leap (5 mins): Moving beyond familiar generative models to “active,” goal-oriented agentic systems.
• Anatomy & Reasoning (12 mins): The core components (brain, memory, tools) and the ReAct “thinking” loop that enables decisions.
• LIVE DEMO (8-10 mins): A live walk-through of a pre-built agent (e.g., in LangChain) that uses two tools (like a web search and a calculator) to answer a complex query.
• The Ecosystem (10 mins): A quick tour of key frameworks (like LangChain) and multi-agent systems (like AutoGen) used to build and orchestrate agents.
• The Frontier & Q&A (5 mins): The biggest research challenges (reliability, long-term planning, security) and open questions.

List of Posters

• Yusri Al-Sanaani: 3D Conditional Image Synthesis of Left Atrial LGE MRI from Composite Semantic Masks
• Ebrahim Ali: Trainable Fractional Spectrogram for Phonocardiogram Classification
• Dhiraj Bhattacharjee: A Scalable Framework for QoE-Driven Dynamic Scheduling and Routing in LEO Satellite Mega-Constellations
• Curtis Davies: Event-Centric Explanations in BDI Agents
• Calvin Jary: A Machine Learning Approach for Next Location Prediction
• Jintao Liang: Green Traffic Engineering for Satellite Network Using Software-Defined Networks and Segment Routing
• Fatemeh Mirshafiee: Detecting Gender Bias to Enhance Inclusivity in Software Engineering Education
• Ali Mohammadi: Prompt Optimization for LLM Code Generation
• Amir Morshedi: Machine Learning-Based Adaptive Clustering for Vehicular Cloud Networks
• Bardia Parmoun: Applications of Autonomous BDI Agents in Robotics
• Saba Pedram: Obtaining Confidence by Construction for Automated Functional Testing
• Maryam Salamatmoghadasi: Integrating Terrestrial and Non-Terrestrial Networks for Sustainable 6G Operations: A Latency-Aware Multi-Tier Cell-Switching Approach
• Mir Md. Saym: Network Slice Admission Control, Resource Allocation, and Migration in Multi Time-slots
• Donya Setyaesh: Exploring surveillance and tracking
• Mahya Shahmohammadimehrjardi: Two-Step Microphone Array Fusion Algorithm for Enhanced Indoor Sound Source Localization
• Andrej Vukovic: A Combined Dynamic Mode Decomposition–Constant False Alarm Rate Algorithm for Near-Real-Time Moving Target Detection
• Ali Zare: The Impact of GitHub Copilot on the Security of Code by Novice and Experienced Developers