CV
408-515-9157 ♦ Atlanta, GA ♦ San Ramon, CA ♦ US Citizen ♦ akash.harapanahalli@gmail.com ♦ akashhara.com ♦ github.com/Akash-Harapanahalli ♦ Google Scholar
Summary
Graduate Research Assistant in the Formal Methods & Autonomous Control of Transportation Systems (FACTS) Lab at Georgia Tech, under the supervision of Dr. Samuel Coogan, currently pursuing a PhD in Machine Learning. Research is centered around safety verification of nonlinear AI-based control systems, mainly through reachability, invariance, and contraction analysis. Developer of immrax, a JAX toolbox used to scalably certify safe performance of neural networks in safety-critical control systems and to train neural networks with robust safety guarantees outright.
Current Position
Georgia Institute of Technology — Atlanta, GA
Machine Learning PhD Candidate, Department of Electrical and Computer Engineering
Aug 2022 - Present
Education
Georgia Institute of Technology — Atlanta, GA
PhD Machine Learning — GPA: 4.00
Aug 2022 - 2027
MS Mathematics — GPA: 4.00
Aug 2022 - Aug 2025
BS Computer Engineering, with Highest Honor — GPA: 4.00
Aug 2019 - May 2022
Technical Skills
ML & Scientific Computing: JAX, PyTorch, Tensorflow, NumPy, SciPy, SymPy, DRAKE, immrax (author)
Math, Control & Robotics: Neural network verification, Reachability analysis, Contraction theory, Lie groups, Differential geometry, Model predictive control
Programming Languages: Python, C/C++, MATLAB, SystemVerilog, Embedded C / RTOS
Tools & Platforms: Linux, Claude Code, Git, ROS2, Arduino, Raspberry Pi, Onshape, Solidworks, Eagle PCB
Research Experience
Graduate Research Assistant — FACTS Lab, Georgia Tech
Aug 2022 - Present
Research is centered around safety verification in AI-based control of nonlinear dynamical systems.
Research Philosophy: Develop rich mathematical theory alongside useful computational tools.
- AI models have little robustness guarantees; this is dangerous when applied in safety-critical domains.
-
Developed a mathematical framework and efficient, GPU-scalable, differentiable toolbox called
immraxfor parametric reachability analysis of nonlinear systems with neural network controllers. - Trained neural network controllers for dynamical systems with formal guarantees of safe operation.
- Developed algorithm for efficient reachable set computation in systems evolving on Lie group state spaces.
- Studied contracting dynamical systems evolving on homogeneous manifolds with Riemannian metrics.
- Mentoring a PhD student and an undergraduate student, building technical communication skills.
Athena Controls Lead — LIDAR Lab, Georgia Tech
Jan 2020 - May 2022
- Led and mentored a subteam of 8 undergraduates to design manipulator planning and control solutions for a humanoid robot using position, velocity, and force approaches.
- Guided peers through a ramping process to understand complex math and background theory.
- Used the DRAKE C++ toolbox to simulate dynamic physical conditions and validate control methods.
- Designed impedance controller for robust, compliant task-space position and force robot manipulation.
Publications [Google Scholar]
* indicates equal contribution
Preprints
- V. Madabushi, A. Harapanahalli, S. Coogan, M. Tucker, “Differentiable Invariant Sets for Hybrid Limit Cycles with Application to Legged Robots.” To appear in IEEE Conference on Decision and Control (CDC), 2026. [pdf]
Journal Articles
- A. Harapanahalli, S. Coogan, A. Davydov, “Learning Certified Neural Network Controllers Using Contraction and Interval Analysis.” IEEE Control Systems Letters (L-CSS), 2026. [pdf]
- A. Harapanahalli, S. Coogan, “A Linear Differential Inclusion for Contraction Analysis to Known Trajectories.” IEEE Transactions on Automatic Control (TAC), 2025. [pdf] [slides]
- B. Gould, A. Harapanahalli, S. Coogan, “Automatic and Scalable Safety Verification Using Interval Reachability With Subspace Sampling.” IEEE Control Systems Letters (L-CSS), 2025. [pdf]
- S. Jafarpour*, A. Harapanahalli*, S. Coogan, “Efficient Interaction-Aware Interval Analysis of Neural Network Feedback Loops.” IEEE Transactions on Automatic Control (TAC), 2024. [pdf]
- A. Harapanahalli, S. Jafarpour, S. Coogan, “Forward Invariance in Neural Network Controlled Systems.” IEEE Control Systems Letters (L-CSS), 2023. [pdf] [slides]
- L. Baird, A. Harapanahalli, S. Coogan, “Interval Signal Temporal Logic From Natural Inclusion Functions.” IEEE Control Systems Letters (L-CSS), 2023. [pdf]
Conference Papers and Abstracts
- A. Harapanahalli, S. Coogan, “Efficient Norm-Based Reachable Sets via Iterative Dynamic Programming.” American Control Conference (ACC), 2026. [pdf]
- A. Harapanahalli, S. Coogan, “Certified Robust Invariant Polytope Training in Neural Controlled ODEs.” Learning for Dynamics and Control Conference (L4DC), 2026. [pdf] [code] [slides]
- A. Harapanahalli, S. Coogan, “Parametric Reachable Sets Via Controlled Dynamical Embeddings.” IEEE Conference on Decision and Control (CDC), 2025. [pdf]
- A. Harapanahalli, S. Coogan, “A Global Coordinate-Free Approach to Invariant Contraction on Homogeneous Manifolds.” American Control Conference (ACC), 2025. [pdf] [slides]
- A. Harapanahalli, S. Coogan, “Efficient Reachable Sets on Lie Groups Using Lie Algebra Monotonicity and Tangent Intervals.” IEEE Conference on Decision and Control (CDC), 2024. Runner-Up for the Hybrid Systems Technical Committee Outstanding Student Paper Prize [pdf] [slides]
- A. Harapanahalli, S. Jafarpour, S. Coogan, “immrax: A Parallelizable and Differentiable Toolbox for Interval Analysis and Mixed Monotone Reachability in JAX.” IFAC Conference on Analysis and Design of Hybrid Systems (ADHS), 2024. [pdf] [github] [slides]
- A. Harapanahalli, S. Jafarpour, S. Coogan, “Contraction-Guided Adaptive Partitioning for Reachability Analysis of Neural Network Controlled Systems.” IEEE Conference on Decision and Control (CDC), 2023. [pdf] [slides]
- A. Harapanahalli, S. Jafarpour, S. Coogan, “A Toolbox for Fast Interval Arithmetic in numpy with an Application to Formal Verification of Neural Network Controlled Systems.” Workshop on Formal Verification of Machine Learning, ICML (WFVML), 2023. [pdf] [github] [poster]
- S. Jafarpour*, A. Harapanahalli*, S. Coogan, “Interval Reachability of Nonlinear Dynamical Systems with Neural Network Controllers.” Learning for Dynamics and Control Conference (L4DC), 2023. [pdf] [poster]
- A. Harapanahalli*, E. Muly*, H. Welch*, T. Brumfiel*, Z. Weng*, M. Akhtar, A. R. Abouelnasr, A. Newland, J. Bunting, K. McGorrey, J. S. Lee, G. Wang, L. Drnach, D. J. Lee, Y. Zhao, “Towards a Biomimetic and Dexterous Robot Avatar: Design, Control, and Kinematics Considerations.” IEEE/ASME International Conference on Advanced Intelligent Mechatronics (AIM), 2020. Best Latest Breaking Results Poster; First Place in the Hardware, Devices & Robotics Track of the 2021 GaTech VIP Innovation Competition [pdf] [abstract] [poster] [video]
Honors and Awards
2025-2026 Herbert P. Haley Fellowship — ML PhD Program, Georgia Tech
Dec 2025
Georgia Tech President’s Fellowship
2022 - 2026
Runner Up for Outstanding Student Paper Prize — IEEE CSS Hybrid Systems TC
2024
ECE Senior Scholar Award (4.00 GPA) — School of ECE, Georgia Tech
Apr 2022
Faculty Honors (every semester) — Georgia Tech
Fall 2019 - Spring 2022
Best Overall Hack — HackGT Horizons 2022
Mar 2022
Best Hardware Hack — HackGT 8
Oct 2021
First Place — 2021 Georgia Tech VIP Innovation Competition (Hardware, Devices & Robotics)
Apr 2021
Best Latest Breaking Results Poster — 2020 IEEE/ASME Intl. Conf. on Adv. Int. Mechatronics
Jul 2020
Best in Show — HackGT Horizons 2020
Feb 2020
Second Place, Best TI Integration — The Invention Studio’s Hack-A-Thing 2020
Jan 2020
State Tournament Champion — 2019 California VEX Robotics High School State Championship
Mar 2019
Academic Service
Student Activities Lead of IEEE CSS Hybrid Systems TC [link]
May 2026 - Present
General Chair of the Decision and Control Laboratory Student Committee [link]
Aug 2025 - Present
Organizer of IEEE CSS Hybrid Systems TC Seminar Series [link]
Jul 2025 - Dec 2025
Chair of 2025 Decision and Control Symposium Organizing Committee [link]
2025
Reviewer for the following venues:
Automatica
2024 - 2025
IEEE Control Systems Letters (L-CSS)
2023 - 2026
IEEE Control Systems Magazine
2026
IEEE Conference on Decision and Control (CDC)
2023 - 2025
American Control Conference (ACC)
2023 - 2025
IEEE Int’l Conference on Intelligent Transportation Systems (ITSC)
2024
Teaching and Professional Experience [Portfolio]
Graduate Teaching Assistant — School of ECE, Georgia Tech
Aug 2022 - Dec 2022
- ECE 3803 — Optimization for Information Systems
- ECE 3084 — Signals and Systems
Applications Engineering Intern — Texas Instruments
Jun 2021 - Aug 2021
- Designed modern Over-The-Air update process for Wi-Fi network processors using embedded C with RTOS.
- Analyzed customer feedback to optimize the design for utility, functionality, and practicality.
- Developed interpersonal professional skills while learning about product lifecycle and post-market strategy.
Other Experiences and Projects [Portfolio]
Data Acquisition Technician — GT Off-Road (BAJA SAE) [github]
Jan 2020 - May 2022
- GT Off-Road designs and manufactures a car to compete in the BAJA SAE competition cycle.
- Rebuilt embedded software infrastructure with a custom dynamic communications protocol using C++ to streamline data transfer in a mesh network of microcontrollers communicating over wired and wireless regimes.
- Designed and programmed a dashboard for a driver to read critical measurements using Solidworks and Eagle.
- Built a closed-loop control algorithm for an Electronic Continuous Variable Transmission (eCVT) in C++.
- Created world’s first fully volumetric high-density holographic persistence of vision display (3D hologram).
- Designed an algorithm to convert arbitrary mesh files to a custom cylindrical 3D display video format.
- Wrote embedded software in C++ and SystemVerilog to encode/decode 3D video over HDMI.
- An interactable piano using stepper motor vibrations to generate music with computer vision features.
- Personally wrote all embedded C++ code to take inputs, schedule, and control stepper motors.
- Repurposed an old 3D printer as a data visualization tool and art piece under strict time constraints.
- Personally designed the system and mounts in Onshape and wrote embedded C++ to control stepper motors.
May 2021 - May 2021
- TOM is a service organization sponsoring open source solutions to the challenges of underserved people.
- Created a low-force oxygen cord reel designed for patients with COPD to avoid tripping hazards.
- Personally helped design/manufacture using Onshape, and wrote embedded C++ to control the device.
- Integrated art and technology by engineering a music instrument out of trashed beer bottles.
- Modeled mounts and connections in Solidworks, wrote embedded Arduino software using MATLAB.
Industrial Screw Sorter — Hack-a-Thing — Second Place, Best TI Integration [github]
Sep 2019 - Jan 2020
- Competed in a five-month-long competition to create an industrial screw sorting apparatus.
- Personally worked on its design and on reusable embedded C++ for its TI MSP-432 using Energia.