Principal Engineer

Vahid Soleimani

Director of Engineering & Principal 3D Computer Vision (SLAM) Engineer

Kudan Computer Vision and Artificial Perception Ltd, Bristol, UK

Senior engineering leader and computer vision researcher with 20+ years at the intersection of academia and industry. I architect SLAM and 3D perception systems that enable machines to autonomously navigate the physical world.

Vahid Soleimani

Background

Principal Computer Vision Engineer with a PhD in Computer Vision from the University of Bristol and over 20 years of combined academic and industrial experience in AI algorithm development and applied research. My expertise centres on Visual and LiDAR SLAM, 3D perception, and real-time sensor fusion for localisation and mapping systems for robotics, AR/VR, and autonomous navigation. I combine rigorous mathematical foundations with hands-on, efficient implementation to deliver robust, production-grade solutions.

As Director of Engineering at Kudan, I lead cross-functional teams in architecting mapping and localisation systems that enable machines to autonomously navigate the physical world. My role bridges technical strategy with delivery—translating cutting-edge research into scalable, industrial-strength products while mentoring engineers and shaping the technical roadmap.

Prior to my industry career, I spent nearly a decade as an academic lecturer at Razi University, teaching topics ranging from fundamental principles to advanced concepts in algorithm design and AI programming. This academic foundation strengthens my ability to communicate complex technical concepts and mentor engineering teams.

20+
Years Experience
7+
Years in SLAM
9+
Years Academic
30+
Publications

Career Journey

Kudan Limited

September 2018 — Present · 7+ years

January 2025 — Present

Director of Engineering & Principal 3D Computer Vision Engineer

Leading engineering teams and driving innovation in SLAM and perception systems for industrial applications.

February 2024 — March 2025

Head of Software Development Team

Managed software development initiatives and coordinated cross-functional engineering efforts.

April 2023 — January 2024

SLAM Team Lead

Led specialised SLAM development team, overseeing technical direction and delivery of core algorithms.

July 2020 — March 2025

Senior 3D Computer Vision (SLAM) R&D Engineer

Advanced R&D on state-of-the-art SLAM systems for industrial applications.

September 2018 — July 2020

3D Computer Vision (SLAM) R&D Engineer

Developed state-of-the-art SLAM systems for industrial applications.

University of Bristol

July 2014 — July 2018 · 4 years

July 2014 — July 2018

PhD Researcher — EPSRC SPHERE Project

Conducted research in video monitoring work package. Also served as Teaching Support Assistant for Image Processing, Computer Vision, C Programming, and Signals courses.

Razi University

February 2005 — July 2014 · 9+ years

February 2005 — July 2014

Lecturer

Taught wide range of computer engineering topics from fundamentals to advanced AI and algorithm design.

Hoorpendar Computer Technology Industries

August 2002 — February 2005 · 2.5 years

February 2003 — February 2005

Senior C++ Embedded Software Engineer

Developed embedded systems for dams, power stations, and industrial facilities including data loggers, turbine controllers, and fibre-optic communication systems.

August 2002 — February 2003

C++ Embedded Software Engineer

Developed data communication platforms and weather monitoring systems for national infrastructure projects.

Academic Background

2014 — 2018

Doctor of Philosophy (PhD) — Computer Science, Computer Vision

University of Bristol, United Kingdom

2002 — 2005

Master of Science (MSc) — Artificial Intelligence and Robotics

Amirkabir University of Technology (Tehran Polytechnic), Iran

1998 — 2002

Bachelor of Engineering (BEng) — Computer Hardware Engineering

National University of Iran (Shahid Beheshti University), Iran

Selected Research Work

Depth-Based Whole Body Photoplethysmography in Remote Pulmonary Function Testing

Introduces a novel depth-based photoplethysmography (dPPG) approach using two opposing RGB-D sensors to reduce motion artefacts in respiratory measurements. The method decouples trunk movements from respiratory motions, significantly improving accuracy of 11 clinical PFT measures compared to single-sensor approaches.

IEEE Transactions on Biomedical Engineering, 2018

View on Scholar →

Remote, Depth-Based Lung Function Assessment

Proposes a remote, non-invasive approach to pulmonary function testing using a depth sensor. By constructing 3D models of the chest and estimating volume variation, the method generates clinical spirometry measures (FVC, FEV1, VC, IC) validated against 85 patients, establishing that chest surface motion is linearly related to lung volume changes.

IEEE Transactions on Biomedical Engineering, 2017

View on Scholar →

Markerless Active Trunk Shape Modelling for Motion Tolerant Remote Respiratory Assessment

Presents a vision-based, motion-tolerant approach for remote lung volume estimation during spirometry tests. Using temporal modelling of trunk shape from two opposing Kinects, the method extracts chest-surface respiratory patterns via PCA on geometrical features, enabling accurate assessment despite subject movement.

IEEE International Conference on Image Processing (ICIP), Athens, 2018

View on Scholar →

3D Data Acquisition and Registration Using Two Opposing Kinects

Presents an open-source data acquisition and calibration system using two opposing Kinect V2 sensors for dynamic 3D object reconstruction. The method estimates relative pose through calibration and registers point clouds at frame level, enabling accurate measurements validated on known-size objects and human subjects.

IEEE International Conference on 3D Vision (3DV), Stanford, 2016

View on Scholar →

Toward Respiratory Assessment Using Depth Measurements from a Time-of-Flight Sensor

Explores using consumer 3D time-of-flight cameras for remote respiratory monitoring. The study validates chest volume estimation during spirometry on 100 patients with various respiratory conditions, demonstrating accurate respiratory rate tracking and FVC estimation within ±1%, with potential for low-cost home monitoring.

Frontiers in Physiology, 2017

View on Scholar →

Improving Ant Colony Optimisation for Brain MRI Image Segmentation and Brain Tumour Diagnosis

Proposes an improved ant colony algorithm for medical image segmentation. The approach incorporates ant direction and tendency when calculating site selection probability, balancing directional influence with pheromone distribution. Applied to brain MRI segmentation and tumour diagnosis with enhanced efficiency.

IEEE Conference on Pattern Recognition and Image Analysis (PRIA), 2013

View on Scholar →

Technical Expertise

SLAM & Localisation

  • Visual SLAM / Odometry
  • LiDAR SLAM / Odometry
  • Sensor Fusion (Camera, IMU, 2D/3D LiDAR)
  • Sensor Fusion (Wheel Odometry, GNSS, INS)
  • Loop Closure / Relocalisation / Place Recognition
  • Bundle Adjustment / Pose Graph Optimisation

3D Perception & Vision

  • Multi-View Geometry
  • Feature Detection & Matching
  • Epipolar Geometry
  • Depth Estimation
  • Camera Calibration
  • 3D Reconstruction

Programming & Frameworks

  • C++ (Modern C++11-20)
  • Eigen / OpenGV
  • Python
  • OpenCV
  • PCL (Point Cloud Library)
  • CMake / Git

Mathematical Foundations

  • Linear Algebra & Geometry
  • Non-linear Optimisation
  • Factor Graph Optimisation
  • Probabilistic State Estimation
  • Kalman Filtering (KF, EKF, ESKF, IEKF, IESKF)

Platforms & Hardware

  • RGB-D Sensors (Kinect, RealSense)
  • LiDAR Systems
  • Stereo Cameras
  • Embedded Systems
  • Real-time Systems

Leadership

  • Technical Strategy
  • Team Leadership
  • R&D Management
  • Technical Roadmap Planning
  • Cross-functional Collaboration

Get In Touch

Interested in collaborating on computer vision projects, discussing SLAM technologies, or exploring research opportunities? I'd love to hear from you.