Nazmus Shakib Sayom

Nazmus Shakib Sayom

PhD Researcher University of Utah

Current Research

Property-Guided Falsification for CPS Control Stacks

Click to enlarge

Developing automated techniques to discover security violations in cyber-physical systems by building property-scoped reductions of control logic and physical dynamics. This research focuses on accelerating violation discovery through targeted analysis of critical system properties, with validation on industry-standard platforms like PX4 and ArduPilot to bridge the gap between theoretical verification and practical autonomous system security.

Figure 1: Flight trajectory analysis demonstrating normal versus low battery emergency descent patterns in simulated drone systems

Safety Policy Verification in Multi-Robot Systems

Investigating the formalization and verification of safety policies in distributed multi-robot environments where multiple stakeholders operate with varying roles, permissions, and operational constraints. The work involves developing formal frameworks for modeling role-based permission constraints, analyzing policy conflicts, and implementing runtime monitoring systems for applications like autonomous vehicle fleets and drone swarms.

Research Focus

3 Active Projects

2024 PhD Started

Falsification

CPS Control Stacks

Verification

Multi-Robot Safety

Reverse Engineering

Binary analysis for CPS

Property-Guided Falsification for CPS Control Stacks

Approach

Developing property-scoped reductions of control logic and physical dynamics to accelerate violation discovery in cyber-physical systems. The research focuses on creating efficient methods that can validate counterexamples on industry-standard platforms.

Platforms

PX4 Autopilot

ArduPilot

SITL Simulation

Hardware-in-the-Loop

Impact

This work aims to bridge the gap between theoretical verification methods and practical autonomous system validation, enabling faster detection of safety-critical vulnerabilities.

Safety Policy Verification in Multi-Robot Systems

Challenge

Formalizing safety policies for multi-stakeholder robot systems where different entities have varying roles, permissions, and constraints that must be verified at runtime.

Methods

Policy Conflict Analysis

Runtime Monitoring

Distributed Consensus

Formal Specification

Applications

The research enables safe coordination in scenarios like autonomous vehicle fleets, drone swarms, and industrial robot networks with multiple operators.

Reverse Engineering

Focus

Vulnerability discovery through binary analysis and reverse engineering, aided by cyber-physical surrogate modeling to guide input generation and state exploration for CPS targets.

Techniques

Static Analysis

Dynamic Instrumentation

Symbolic Execution

Firmware Emulation

Protocol Analysis

Tooling

Ghidra

QEMU / Unicorn for emulation

AFL++/LibFuzzer with model-guided inputs

Career Timeline

Aug 2024 – Present

PhD in Computer Science

University of Utah

Conducting doctoral research on cyber-physical systems security, developing falsification techniques and safety verification methods for autonomous systems.

Research CPS Security Formal Methods

Jan 2023 – Jul 2024

Linux Developer

meldCX, Dhaka

Developed application-specific embedded Linux distributions for ARM platforms. Implemented secure, network-accessible abstractions for device drivers.

Linux ARM Embedded Security

Jan 2022 – Dec 2022

Lecturer

Daffodil International University

Taught Operating Systems, Embedded Systems, and Computer Fundamentals. Co-authored conference papers in deep learning and secured research funding.

Teaching Research Deep Learning

Apr 2021 – Dec 2021

Senior Software Engineer

HelloTask Platform Ltd.

Led design and development of jobs platform for blue-collar workers. Built IVR broadcasting system for users without internet access.

Leadership Platform Development Accessibility

Jun 2020 – Mar 2021

Embedded Systems Engineer

HelloTask Platform Ltd.

Prototyped self-checkout store network and centralized advertising delivery platform using IoT and computer vision analytics.

IoT Computer Vision Prototyping

2017 – 2020

B.Sc. Computer Science & Engineering

American International University-Bangladesh

Graduated Summa Cum Laude with CGPA 3.96/4.00. Completed thesis on disaster victim tracking and rescue support systems.

Summa Cum Laude Emergency Systems IoT

Project Portfolio

Click to enlarge

Computer Vision API

OpenGaze

RESTful web API service for OpenFace facial behavior analysis, providing eye-gaze and head-pose data for research applications.

FastAPI OpenCV Computer Vision

Click to enlarge

Embedded Systems

DePen

Raspberry Pi Zero powered optical text scanner using computer vision for OCR operations and word definition lookup.

Raspberry Pi OpenCV Tesseract

Click to enlarge

RF Communication

TALK-E

Long-range multi-channel walkie-talkie system with custom pre-amplification for emergency communication scenarios.

Arduino nRF24L01 RF

Click to enlarge

Data Visualization

Dancing Bees

Python-based trajectory visualization tool for analyzing honey-bee flight patterns from annotated video frames.

Python OpenCV Visualization

Undergraduate Thesis

Disaster Victim Tracking System

Comprehensive emergency response system utilizing multi-layer communications including ESP32, LoRaWAN, BLE, and GSM for resilient search-and-rescue operations.

IoT LoRaWAN Emergency Systems Multi-layer Communication

Let's Connect

Available for research collaborations

Email

sayom.shakib@utah.edu

Research discussions and collaborations

GitHub

github.com/nsssayom

Open source projects and contributions

Google Scholar

View Publications

Publications and citations

LinkedIn

linkedin.com/in/nsssayom

Professional network and career updates

Open to Collaboration

Open to collaboration on CPS Security and Safety, Reverse Engineering and related works. Also open for Summer Internship opportunities in 2026. Feel free to reach out!