Harrish Joseph

PhD Researcher bridging the frontiers of Scientific Machine Learning, Spacecraft Dynamics, and Complex Systems

Connect on LinkedIn

I'm a PhD researcher at Sapienza University of Rome, specializing in Scientific Machine Learning for nonlinear dynamical systems. My work focuses on developing novel deep learning architectures—Neural Operators, Neural ODEs, and Temporal Fusion Transformers—for parameter identification and structural health monitoring.

With a background in Aerospace Engineering and space missions, I bring an interdisciplinary perspective to understanding complex systems, from spacecraft dynamics to evolutionary biology, always asking the fundamental questions about emergence, chaos, and the meaning encoded in dynamic patterns.

Journey Through Space & Time

🧠
2022-Present

PhD Research

Scientific Machine Learning for nonlinear systems. Developed Temporal Fusion Transformer, Neural ODEs, and Operator Learning for damage detection and system identification.

🔬
2022

Research Engineer

Optimized Bouc-Wen parameters achieving 25% error reduction. Developed deep neural networks for metamaterial vibration prediction with 90% accuracy.

🚀
2017-2021

Masters in Space Engineering

Sapienza University of Rome. Focus on Space Flight Mechanics, Spacecraft Design, Space Robotics. Thesis: Range Delay in JUICE mission tracking.

🎓
2010-2014

Bachelor of Technology

Aerospace Engineering at Karunya University. Mathematical modeling of wave propagation in thin-walled structures.

Research Impact

🎯
0%
Error Reduction
0%
Prediction Accuracy
🚀
0+
Research Projects

Expertise Areas

🧠

Scientific ML

Neural Operators, Physics-Informed Neural Networks, Autoencoders, GANs, and Temporal Fusion Transformers for dynamical system analysis.

🚀

Space Systems

Spacecraft dynamics, orbital mechanics, mission analysis, and space navigation systems. Experience with JUICE mission and AGI STK.

🌊

Nonlinear Dynamics

Parameter identification, structural health monitoring, damage detection in complex hysteretic systems like Duffing oscillators.

🤖

Robotics & Control

Dynamic modeling, trajectory planning, Extended Kalman Filtering, and closed-loop control for manipulators and autonomous systems.

Skills Constellation

Machine Learning

TensorFlowPyTorchJAXKerasScikit-Learn

Programming

PythonMATLABJavaScriptBash

Scientific Computing

NumPySciPyPandasMatplotlib

Aerospace Tools

AGI STKNAIF-SPICEMSC Adams

Engineering Design

CATIASolidWorksInventorHypermesh

Version Control

GitGitHub

Publications

Deep learning architectures for data-driven damage detection in nonlinear dynamic systems

H. Joseph, G. Quaranta, B. Carboni, and W. Lacarbonara

Nonlinear Dynamics, vol. 112, no. 23, pp. 20 611–20 636, Dec. 2024

View Publication →

+ Conference proceedings and collaborative research papers

Let's Explore the Cosmos Together

Interested in collaborating on research at the intersection of machine learning, dynamical systems, and space exploration? Let's connect and push the boundaries of what's possible.

Connect on LinkedIn