AI, Robotics, Networking, and IoT
AI & Machine Learning Engineer · Researcher
Engineering AI and Machine Learning Systems for the Real World.
I build AI and Machine Learning systems that move from research to reliable deployment. My work spans predictive robotics, low latency networking, computer vision, cybersecurity analytics, and industrial edge platforms.
Remote Robotic Surgery Studies
BBRv2 TI Testbed Results
Healthcare Security Analytics
01 / Current Work
Hitachi Digital Services · Toyota Vision Inspection Platform
AI Enabled Industrial Vision Supported by Dependable Edge Software.
I engineer edge services across the inspection pipeline from industrial camera ingestion and image storage to model integration, structured results, MQTT messaging, container delivery, and fleet orchestration. The work connects software reliability with computer vision in a production manufacturing environment.
01AI and Vision IntegrationImages · Models · Results
02Industrial MessagingMQTT · NanoMQ · Event Flow
03Edge Application DeliveryContainers · CI/CD · ZEDEDA
04Cloud and Data IntegrationAWS ECR · Storage · Observability
02 / Research Direction
Ongoing Research01 02 03
Advancing Predictive AI for Vision Assisted Teleoperation.
I am continuing this research by bringing together future tool motion and scene prediction, adaptive prediction horizons for 50 to 300 milliseconds of video delay, calibrated confidence, predictive overlays, and reproducible robotic testbeds.
As a researcher specializing in AI and Machine Learning, my goal is to build systems that remain useful when communication is delayed, observations are incomplete, and decisions are time critical. The system should communicate what it predicts, how certain that prediction is, and when the human operator should remain in control.
Vision Assisted Teleoperation
Predictive visual guidance that helps an operator understand likely tool and scene motion during network delay.
Uncertainty Aware Prediction
Confidence estimates and adaptive horizons that make model limitations visible instead of hiding them.
Reproducible Edge Intelligence
Testbeds that connect Machine Learning, communication behavior, edge computing, and measurable system outcomes.
03 / Profile
I connect AI and Machine Learning research with deployable edge and Industrial IoT systems.
My work begins with physical systems, imperfect data, network delay, and deployment constraints. I develop the prediction, estimation, and communication methods needed to make those systems more reliable, then translate the research into practical software and evaluation workflows.
At Hitachi Digital Services, I contribute to the Toyota Vision Inspection Platform, where computer vision, industrial cameras, edge services, messaging, storage, and containerized deployment must operate as one dependable inspection workflow.
At Iowa State University, my research investigates Tactile Internet enabled Remote Robotic Surgery. I study how Informer models, system identification, Kalman filtering, and QUIC congestion control can reduce the effect of delay, jitter, and packet loss on remote operation.
04 / Experience
Engineering, Research, Cybersecurity, and Teaching.
A professional record built across deployed systems, peer reviewed research, healthcare security, and engineering education.
March 2026 to PresentToyota Vision Inspection Platform
Computer Edge Developer
Hitachi Digital Services
Developing and integrating edge services for an automotive vision inspection platform. The work connects industrial cameras, image storage, machine learning inference, structured results, device messaging, and repeatable software delivery across distributed edge environments.
May 2025 to October 2025Clinical Network Analytics
Machine Learning Analyst, Cybersecurity
Mary Greeley Medical Center
Built Python and PySpark pipelines that transformed Zeek and Suricata telemetry into enriched security features. Deployed Malcolm and Hedgehog sensors, prototyped anomaly detection in OpenSearch, and developed a local RAG assistant. The resulting workflow reduced false positives by 30% and improved triage time by 40%.
August 2022 to December 2025Real Time Intelligent Systems
Graduate Research Assistant
Iowa State University
Developed predictive estimation and communication methods for delay sensitive robotic systems. Informer based studies achieved approximately 95% position prediction accuracy, while BBRv2 TI experiments reduced jitter by 98% and improved throughput 1.5 times relative to the baseline.
January 2019 to July 2022Teaching and Project Supervision
Faculty Lecturer
Quaid e Awam University of Engineering, Science and Technology
Taught Python, C++, embedded systems, digital logic, computer networks, and electronics. Supervised undergraduate work in Machine Learning, IoT, automation, and control systems while managing practical laboratory instruction.
SabBachao
Developed Python based pricing and comparison methods for grocery cost optimization, including data collection, API integration, and backend design.
Poulta
Led an IoT environment monitoring system using ESP32, LoRa, and GSM to stream sensor data and support practical farm analytics.
05 / Research
Active Research Program
Predictive AI and Communication for Tactile Internet Enabled Remote Robotic Surgery.
My research investigates how intelligent prediction, state estimation, and low latency communication can improve reliability when delay, jitter, and packet loss affect remote robotic systems. Predictive studies achieved approximately 95% position accuracy, while network experiments reduced jitter by 98% and improved throughput 1.5 times relative to the baseline. I continue to extend this work toward vision assisted teleoperation and uncertainty aware decision support.
Predictive AI
Informer based forecasting for future robotic tool position during burst packet loss and changing network conditions.
State Estimation
MOESP system identification combined with Kalman filtering for data driven position estimation.
Tactile Internet Transport
Utility guided BBRv2 TI congestion control over QUIC, balancing throughput, delay, jitter, and loss.
Vision Assisted Operation
Ongoing work on predictive overlays, confidence, adaptive horizons, and reproducible robotic evaluation.
Selected Peer Reviewed Publications
Complete Google Scholar Profile2026IEEE CCNCBBRv2-TI: A Utility-Guided Congestion Control Algorithm for Tactile Internet over QUICFirst Author · Conference Paper2025IEEE Transactions on Medical Robotics and BionicsEnhanced Position Estimation in Tactile Internet-Enabled Remote Robotic Surgery Using MOESP-Based Kalman FilterFirst Author · Journal Article2025SensorsA Predictive Approach for Enhancing Accuracy in Remote Robotic Surgery Using Informer ModelFirst Author · Journal Article2024IEEE AccessEnergy-Efficient Framework to Mitigate Denial of Sleep Attacks in Wireless Body Area NetworksCoauthor · Journal Article2024IEEE IWCMCEnhancing Precision in Tactile Internet-Enabled Remote Robotic Surgery: Kalman Filter ApproachFirst Author · Conference Paper2023PeerJ Computer ScienceInternet of Things-Based Sustainable Environment Management for Large Indoor FacilitiesFirst Author · Journal Article2018IEEE IoTaISIoT Based Poultry Environment Monitoring SystemFirst Author · Conference Paper
06 / Projects
Selected Work Across AI, Robotics, Security, and Industrial Systems.
Current industry work, ongoing research, and documented projects across AI, Machine Learning, Remote Robotic Surgery, cybersecurity, IoT, and Industrial IoT.
01Industrial Edge AI
Toyota Vision Inspection Platform
Edge software for automotive visual inspection, including camera and image workflows, model integration, message transport, structured inference results, container delivery, and fleet ready deployment practices.
Python · Computer Vision · MQTT · Docker · Kubernetes · AWS · ZEDEDA
Current Private Industry Work02Ongoing Research
Vision Assisted Remote Robotic Surgery
An active research direction that combines future tool motion and scene prediction, adaptive prediction horizons, confidence aware overlays, and reproducible experiments for teleoperation under communication delay.
AI · Machine Learning · Computer Vision · Time Series · Edge Intelligence · Tactile Internet
Research in Progress03Predictive Robotics
Informer for Surgical Time Series
A forecasting pipeline for robotic tool position under burst packet loss, evaluated with JIGSAWS kinematics and a four state network loss model.
PyTorch · Informer · JIGSAWS · Time Series Evaluation
View Repository 04Healthcare Cybersecurity
Real Time Network Anomaly Detection
A telemetry and anomaly detection workflow built around Malcolm, Hedgehog Linux, Zeek, and OpenSearch for security monitoring and investigation.
Zeek · Suricata · OpenSearch ML · Python · Network Telemetry
View Repository 05IoT and Industrial IoT
Poultry Environment Monitoring System
A connected monitoring system using ESP32, LoRa, and GSM to collect environmental measurements and provide practical farm analytics for large indoor facilities.
ESP32 · LoRa · GSM · Sensors · IoT Data Pipelines
View Published Work 06Private Generative AI
Local AI Knowledge Assistant
A local document assistant that ingests PDFs and text, builds a vector index, and supports grounded question answering without sending documents to a cloud model.
Streamlit · LangChain · Ollama · RAG · Vector Search
View Repository 07 / Credentials
Ph.D. in Computer Engineering
Research in predictive estimation and protocol optimization for Tactile Internet enabled Remote Robotic Surgery, including Machine Learning forecasting, state estimation, and low latency communication.
M.E. in Electrical Engineering
Thesis work in IoT based poultry environment monitoring, embedded sensing, and data pipelines. GPA 3.33 of 4.0.
B.E. in Electronics Engineering
Capstone work in earthquake monitoring through wireless sensor networks. GPA 3.86 of 4.0.
Leadership and Service
- 2025 to 2026Graduate and Professional Student Senate Treasurer
- 2024 to 2025Graduate and Professional Student Senate Senator, ECpE
- 2023 to 2024Pakistan Student Association New Student Coordinator
Technical Range
Python, PyTorch, TensorFlow, scikit learn, time series forecasting, anomaly detection, computer vision, YOLO, RAG
Remote Robotic Surgery, Tactile Internet, system identification, MOESP, N4SID, Kalman filtering, uncertainty estimation
Python services, MQTT, NanoMQ, ESP32, LoRa, industrial cameras, Docker, Kubernetes, ZEDEDA, Linux
QUIC, BBRv2 TI, tc, netem, AWS, GCP, OpenSearch, PySpark, pandas, PostgreSQL, CI/CD
Beyond the Work
Cricket, cycling, gym training, nature walks, long drives, repairing laptops and gadgets, and building custom computers.
Languages: English · Urdu · Sindhi