Artificial Intelligence on Mohammad Movahedi

Agentic AI as a Research Tool

Sat, 30 May 2026 00:00:00 +0000

This scratchpad explores how agentic AI can be used as a practical research tool. It focuses on understanding how AI agents can support literature review, data analysis, writing, coding, workflow automation, and knowledge organization.

Local LLMs

Thu, 28 May 2026 00:00:00 +0000

This scratchpad collects my notes on running and working with local large language models. It is meant to better understand model deployment, inference tools, hardware requirements, quantization, fine-tuning, evaluation, and practical workflows for using LLMs outside cloud-based platforms.

Persona-Based Hurricane Evacuation Travel Demand Analysis

Fri, 15 May 2026 00:00:00 +0000

🌀 The Challenge of Rural Evacuation Planning

Florida continues to be the most hurricane-prone state in the United States. While evacuation orders play a significant role in reducing casualties, their effectiveness relies heavily on household decision-making. In rural communities, such as the Florida Panhandle, this is complicated by limited infrastructure, longer travel distances, and resource constraints. Current evacuation demand models often rely on oversimplified assumptions of rational decision-making, failing to capture the stress, urgency, and irrationality inherent in disaster scenarios.

Introduction to LLM

Fri, 01 May 2026 00:00:00 +0000

This scratchpad collects the LLM introduction series, an eight-part walkthrough from transformer basics to responsible deployment. Many examples use evacuation and transportation planning as a running case, but the concepts apply broadly.

Simulating Community Behaviors with LLMs

Thu, 16 Apr 2026 00:00:00 +0000

🚧 The Challenge of Debris Management

Following catastrophic events like Hurricane Ian, post-disaster debris management becomes a critical, time-sensitive logistical challenge. Traditional planning relies heavily on volume estimation and routing, treating communities as passive entities. However, emergent human behaviors—such as illegal debris dumping—introduce highly stochastic burdens that derail recovery efficiency, amplify health risks, and drastically increase municipal costs.

Generative AI + Virtual Reality: A training tool

Tue, 08 Jul 2025 00:00:00 +0000

Sometime in late 2023, when the generative AI community suddenly shifted its focus to picture and video generation, I started developing a framework to first generate textures of materials based on the description that the inspection guidelines provide. This concept quickly evolves into a framework to generate auditory and visual cues for 3D objects.

The proposed framework

In a conference paper titled “Generative Artificial Intelligence and Virtual Reality: Emerging Future of the Building Component Inspection Training”, presented in CIB WBC CIB 2025, I described this framework in detail. The framework integrates generative artificial intelligence (Gen AI) with virtual reality (VR) to create a more immersive and effective training environment for building maintenance and operations. Traditional training methods often rely on static visuals and text, limiting inspectors’ exposure to the diverse and complex conditions found in real-world scenarios. In contrast, the proposed approach leverages Gen AI to generate realistic textures of deteriorating building components, which are then embedded into an interactive VR environment. This enables trainees to engage in hands-on, immersive practice, improving their ability to identify and assess conditions accurately. The framework aims to reduce subjectivity in inspections and enhance knowledge transfer. A hypothetical case study demonstrates the framework’s potential, highlighting its broader applicability not only in building maintenance but also in other domains that require precise, objective evaluations.

Advanced Estimation Models for Demolition Waste

Wed, 20 Mar 2024 00:00:00 +0000

🏗️ The Demolition Data Gap

In the United States, construction and demolition waste (CDW) accounts for roughly 67% of the total solid waste stream, with a staggering 90% originating directly from demolition sites. For a transition toward a circular economy, accurately estimating the recoverable portion of this waste is crucial. Yet, traditional estimation models remain heavily focused on total volume rather than recyclable potential.