Profile banner
Harsimret Singh Dhami

Harsimret Singh Dhami

Space engineer building onboard intelligence for satellite constellations

harsimretdhami
Glasgow, United Kingdom
Joined May 2026
Favicon imageharsimretdhami.com
Favicon imageresearchgate.net

Network

5K connections
🎓
Cranfield Students Alumni
🤖
AI Robotics Engineers
⚙️
Space Systems Engineers
👨‍🏫
Cranfield Faculty Leadership
💨
Aerodynamics Specialists
🚀
Aerospace Propulsion Engineers
✈️
Flight Dynamics Control

Summary

Focus on onboard intelligence and federated learning for satellites: Dhami's portfolio and mission brief emphasise designing federated learning architectures so satellites can train locally and share model updates rather than raw telemetry, enabling real-time anomaly detection, debris avoidance and autonomous in-orbit decision-making. harsimretdhami
Academic work in vision-based GNC and semantic segmentation: As a Cranfield research intern and co-author on IFAC and ACM papers, Dhami has contributed to vision-based guidance, navigation and control architectures and semantic-segmentation mapping systems for safe landing and autonomous drone delivery. ac+1
Hands-on astrodynamics and mission analysis experience from internships: Multiple internships (SmallSpark, SuperSharp) and a role doing mission analysis/flight dynamics show applied experience implementing force models, orbit propagation tools (Orekit), manoeuvre planning and constellation design. harsimretdhami+1
Leadership and community engagement within student and competition settings: Held team leadership and representative roles (Project Manager for a satellite design competition and Course & Careers Representative at Cranfield), indicating experience coordinating multidisciplinary teams and industry engagement. harsimretdhami

Work

Education

Projects

Writing

SSS: Towards Autonomous Drone Delivery to Your Door Over House-Aware Semantics

January 1, 2024

Paper on structural semantic segmentation (SSS) for autonomous drone delivery, proposing methods to identify safe drop-off points for last-hundred-feet delivery using computer vision.

Favicon imagedl.acm.org

MegaSat: Feasibility of a Modular and Expandable Geostationary Platform for Optimizing Geo-Slot Allocations

January 1, 2024

Conference paper exploring a modular, expandable geostationary platform concept to optimise geo-slot allocations and platform scalability.

Favicon imageproceedings.com

Semantic segmentation-based mapping systems for the safe and precise landing of flying vehicles

January 1, 2022

Describes a semantic segmentation-based mapping system to support safe and precise landing of flying vehicles; includes methods and evaluation for vision-based landing site detection.

Favicon imagedspace.lib.cranfield.ac.uk

Hobbies

Building rockets, CubeSats and autonomous systems. harsimretdhami