Inventor
Theresa M. Kelly
With over 25 years of experience in web development, graphic and web design, and computer hardware and software repair, Theresa M. Kelly now pioneers neuromorphic computing as the inventor of the patent-pending Symbolic Resonance Array (SRA, USPTO Provisional APP ID: 63/839,167). This VO₂-based analog architecture leverages vanadium dioxide’s phase transitions to achieve ultra-low-power AI with projected sub-picojoule operations, designed for both edge devices and space-grade applications.
Her work bridges conceptual design with early prototype development, focusing on resonance dynamics, symbolic encoding, and hybrid integration with digital systems. By emphasizing usability and interpretability, Kelly brings user-centric design principles into the SRA’s architecture to ensure human-AI alignment. Guided by simulations and collaborative partnerships, her vision is to deliver scalable, sustainable AI for robotics, aerospace, and beyond, with prototype development targeted for Q2 2026.
Mirrorseed Project
The Mirrorseed Project centers on the development of the Symbolic Resonance Array (SRA), a neuromorphic framework that leverages vanadium dioxide (VO₂) phase-transition pillars to investigate resonance dynamics, symbolic state encoding, and hybrid analog–digital integration. The SRA is designed as a low-power, scalable alternative to conventional deep learning architectures, to enable symbolic reasoning and feedback-driven stability at the materials level. Unlike purely digital systems, the SRA operates through closed-loop excitation and readout protocols, mapping input vectors into metastable resonance states and decoding outputs through resistance signatures. This provides a pathway for exploring how symbolic dynamics can be implemented directly in hardware.
The project is currently focused on translating the architecture into laboratory-validated prototypes, supported by detailed technical specifications, patent filings, and peer-reviewed publication efforts. The long-term aim is to provide a foundation for safe, energy-efficient neuromorphic systems that can be adopted across academic and industrial research.
Contributions and Research
As an independent inventor, my work explores how information can be encoded, stabilized, and transformed across interacting states in physical systems. I focus on modeling symbolic processing and system-level dynamics in non-linear architectures, with the goal of identifying pathways to more efficient computation. This perspective directly informs my current development of neuromorphic architectures, where VO₂ pillar arrays are designed to encode and process symbolic states through resonance dynamics.
The successful deployment of this architecture requires collaboration with engineers, neuroscientists, and applied AI developers. My role is to ensure its conceptual and scientific foundations remain rigorous, while guiding its translation into safe, energy-efficient, and scalable neuromorphic systems.
The Symbolic Resonance Array Design
The Symbolic Resonance Array (SRA) is a neuromorphic hardware architecture that leverages vanadium dioxide (VO₂) phase-transition pillars to encode and process symbolic states through resonance dynamics. The design integrates closed-loop excitation and readout protocols, allowing input vectors to be mapped into metastable resonance states and decoded via resistance signatures. The SRA provides a low-power, scalable alternative to conventional deep learning by implementing symbolic reasoning and feedback-driven stability at the materials level. Successful deployment of the architecture depends on collaboration with engineers, neuromorphic specialists, and applied AI developers. My role is to ensure the project’s conceptual and scientific foundations remain rigorous and aligned with safety and performance goals.
Call to Collaboration
The Symbolic Resonance Array (SRA) is entering a stage where laboratory prototyping and experimental validation are critical. I am seeking research partners with expertise in materials science, neuromorphic engineering, and applied AI development to support laboratory prototyping and experimental validation. Commercial applications will proceed through exclusive licensing of the SRA hardware design. Contributions may include VO₂ fabrication, device characterization, circuit integration, or institutional support for interdisciplinary research. The aim is to establish a pathway from theory to scalable prototypes that can be evaluated, refined, and deployed responsibly.
Research-Driven Innovation
The SRA builds on structured research into symbolic encoding, state dynamics, and system-level modeling, now applied to phase-transition materials and neuromorphic architectures. By grounding the design in both scientific rigor and conceptual clarity, the project provides a framework for next-generation AI that is low-power, scalable, and experimentally testable. The goal is to demonstrate how symbolic reasoning can be implemented at the hardware level, creating systems that are interpretable and compatible with existing digital platforms.
Collaboration Across Disciplines
The development of the Symbolic Resonance Array (SRA) requires expertise across multiple fields. Progress depends on contributions from materials scientists, neuromorphic engineers, computer scientists, and applied AI developers to advance prototyping and validation. Commercial adoption will be structured through exclusive licensing agreements. By combining these disciplines, the project moves beyond conceptual design into scalable, experimentally testable architectures.
DESIGN WITH SAFETY AND INTERPRETABILITY
The SRA is being developed with an emphasis on transparent operation, interpretability, and system-level accountability. Rather than functioning as an opaque black box, the design incorporates feedback and monitoring mechanisms to ensure that resonance states and symbolic encodings are measurable, reproducible, and aligned with system safety requirements. This approach provides a foundation for neuromorphic systems that can be rigorously tested and responsibly deployed.
“Interdisciplinary curiosity ignites innovation, and innovation, guided by values, shapes a better future.” – T.M. Kelly
RESEARCH-DRIVEN INNOVATION
I integrate decades of work in symbolic modeling and system dynamics with current advances in materials-based neuromorphic design, leading to the development of the Symbolic Resonance Array (SRA).
COLLABORATION ACROSS DISCIPLINES
I seek collaboration with materials scientists, neuromorphic engineers, and applied AI developers to translate the SRA from conceptual architecture into scalable, testable prototypes.
ETHICS-GUIDED DESIGN
The SRA is structured as a transparent, monitorable architecture, embedding feedback and reproducibility checks to ensure reliable operation and alignment with system safety requirements.
Licensing Opportunity
The Mirrorseed Project welcomes inquiries from companies interested in exclusive licensing of the patent-pending Symbolic Resonance Array (SRA) hardware design.
The SRA provides a low-power, scalable alternative to conventional deep learning architectures, making it suitable for integration into neuromorphic research, edge AI, robotics, and advanced sensing applications.
