I am a PhD candidate at MIT's Space Telecommunications, Astronomy and Radiation Laboratory with PI Prof. Kerri Cahoy. I work on space robotics, autonomous systems, instrumentation, and machine learning applied to spacecraft. Currently I work on dynamic tasking formulations for Earth-observing satellites funded by Planet Labs so we can use existing satellites more effectively through improved perception, onboard planning, and edge computing. I specifically focus on safe and verifiable autonomy formulations that can be operationalized. I have also conducted research on reinforcement learning for electromagnetic formation flying spacecraft, vision models for remote sensing, and utilizing liquid lenses for laser communications.
I currently work at Planet Labs as an Edge Compute/Machine Learning Engineer. I have worked on Pelican's onboard compute system on the NVIDIA Jetson platform, as the first employee on the Edge Compute team. My previous experience includes integrated photonics at X, the moonshot factory (formerly Google [x]), and I also led all onboard electronics development for Imperial College London Rocketry.
I started this blog for technical sewing projects. There are countless blogs from the early to mid 2010s that have been invaluable to my sewing journey, and I hope to contribute to that body of knowledge. I might occasionally write about fashion, research, or other topics also.
I wrote this blog platform called Spoingo. I also host a tree-based productivity software called Treetrack.
Contact: shreeyam [at] mit [dot] edu
Education
Massachusetts Institute of Technology (MIT)
PhD Spacecraft Systems and Sensors, minor in AI and Computer Vision
2024 - 2025Thesis: Spacecraft Autonomy through Computer Vision and Onboard Planning
SM Aeronautics and Astronautics
2020 - 2022GPA: 5.0/5.0
Thesis: Optical Performance and Prototyping of a Liquid Lens Laser Communications Transceiver
Imperial College London
2016 - 2020MEng Aeronautical Engineering with a Year Abroad
Degree Classification: First Class Honours
Thesis: Design of low leakage MEMS valves for spacecraft applications
Experience
MIT Space Telecommunications, Astronomy and Radiation Laboratory
2020 - PresentResearch Assistant (PI: Prof. Kerri Cahoy)
Cambridge, MA- Trained policies through reinforcement learning for multi-agent electromagnetic formation flight, for use in space as a set of collaborative agents
- Developed analytic and reinforcement learned heuristics for dynamic task rescheduling for agile Earth-observing spacecraft
- Trained machine learning models and created dataset for detection of ocean fronts from satellite imagery
- Fine-tuned vision-language models for spatial visual question answering on remote sensing imagery
- Designed, trained, tested, and validated vision models for cloud segmentation onboard ESA’s OPS-SAT mission
- Conducted space environment testing and prototyping with NASA for a novel lasercom pointing and tracking system using liquid lenses, including testing in microgravity on a parabolic flight
Planet Labs
2022 - 2024Edge Compute/Machine Learning Engineer
San Francisco, CA- Architected next-generation onboard compute platform for Planet’s Earth-observing satellite missions using NVIDIA GPUs as the first hire on the edge compute team
- Designed, specified, and developed hardware for a low-power computer vision instrument for onboard autonomy
- Wrote and deployed code for autonomous operations stack, including Linux camera drivers, model inference, security infrastructure, and operation with main flight computer
- Radiation tested components including NVIDIA Jetson devices at UC Davis cyclotron to qualify hardware for space operations
- Trained and distilled vision models for utility estimation for upcoming imaging activities, to be used for autonomous operations
- Designed algorithms for onboard autonomous task re-scheduling, data collection and deployment campaign in progress
X Development LLC (formerly Google[x])
2021Intern @X
Mountain View, CA- Interned under Taara, aiming to provide low-cost free-space optical communications (FSOC) internet access for underdeveloped countries.
- Conducted communications architecture analysis for optically preamplified direct detection and coherent detection techniques.
- Modeled integrated photonics components to assess the capabilities of each architecture.
Intelligent Environments Europe Ltd (ieDigital)
2016Software Developer Intern
London, UK- Worked as a full stack software developer intern writing C# across teams managed under agile methodology.
- Automated part of the workflow for business analysts using Python, saving tens of hours on importing old functional specifications into JIRA.
Skills
Programming Languages
Python, C, C++, C#, MATLAB, Verilog
Software
Microsoft Office, SolidWorks, Altium, PyTorch, Docker
Languages
English (Native), Hindi (Fluent), Korean (Intermediate)
Projects & Extracurricular Activities
Imperial College London Rocketry
June 2019 - July 2020Electronics & Payload Team Lead
- One of four executive leads of an 80-member team.
- Led design and manufacture of all rocket electronics, including avionics systems, telemetry, data acquisition, throttle control system, and payload.
Awards
- Best Paper Award in Small Satellites at AIAA SciTech Forum (2022)
- Imperial College London Dean's List (2019)
Publications
2025
Liquid lenses for aerospace beam steering and communications: MOSAIC
Shreeyam Kacker, Kerri Cahoy
Optica Optics Express, Vol.33(1)
Leveraging Realtime Meteorological Data for Dynamic Tasking of Agile Earth-Observing Satellites
Shreeyam Kacker, Steve Chien, Kiruthika Devaraj, Işil Demir, Kerri Cahoy
International Workshop on Planning & Scheduling for Space (IWPSS) 2025
Vision-Based Dynamic Tasking for Earth-Observing Satellite Constellations
Shreeyam Kacker, Kiruthika Devaraj, Işil Demir, Steve Chien, Kerri Cahoy
39th Annual AIAA/USU Conference on Small Satellites (forthcoming)
Deep Reinforcement Learning for Multi-Agent Spacecraft Electromagnetic Formation Flight
J Biberstein, Shreeyam Kacker, Kerri Cahoy, S Karaman
in prepation
Vision Language Models for Single Shot Scene Understanding and Dynamic Tasking of Agile Earth-Observing Satellites
Shreeyam Kacker, Kerri Cahoy
in prepation
2024
BeaverCube II: Using AI-Optimized Processors on Earth-Observing CubeSats for Autonomous Image Analysis and Intelligent Data Handling
Adam Bahlous-Boldi, Celvi Lisy, Neelambar Mondal, Brianna Ferro, Shreeyam Kacker, Mary Dahl, Madeline Anderson, Kerri Cahoy
38th Annual AIAA/USU Conference on Small Satellites
Reinforcement-Learned Lookahead Heuristics for Earth-Observing Satellites
Shreeyam Kacker, Kerri Cahoy
38th Annual AIAA/USU Conference on Small Satellites
2023
MOEMS-based lens-assisted beam steering for free-space optical communications
Daniel A Goldman, Paul Serra, Shreeyam Kacker, Lucas Benney, Daniel Vresilovic, Steven J Spector, Kerri Cahoy, Jordan S Wachs
IEEE Journal of Lightwave Technology, Vol.41(9)
Systems and Methods for Cloud Avoidance
Kiruthika Devaraj, Shreeyam Kacker
United States Patent and Trademark Office Application 18/345,883
Fast ocean front detection using deep learning edge detection models
Violet Felt, Shreeyam Kacker, Joe Kusters, John Pendergrast, Kerri Cahoy
IEEE Transactions on Geoscience and Remote Sensing 61
Optical Performance of Commercial Liquid Lens Assemblies in Microgravity
Shreeyam Kacker, Kerri L Cahoy
SPIE Optical Engineering, Vol.62(11)