Summary
Taylor Johnson is an academic leader and researcher specializing in ensuring the safety, security, and trustworthiness of autonomous cyber-physical systems that integrate machine learning, with 11+ years of experience bridging formal methods and software engineering. As Associate Dean for Graduate Education and Director of VeriVITAL at Vanderbilt, they lead a sustained, well-funded research program focused on provable robustness of neural networks and end-to-end safety guarantees for perception, planning, and control in autonomy. Their work spans aerospace, automotive, robotics, power systems, and medical devices and has attracted support from DARPA, NSF, AFRL, NSA, NVIDIA, Toyota and others. Taylor combines deep theoretical expertise (formal verification, control theory, hybrid systems) with practical systems experience (embedded/real-time systems, cloud/edge, FPGA), mentoring on average nine PhD students and guiding translational research through a university-affiliated consultancy and a healthcare startup. They are an active educator with courses on ML verification, automated verification, and embedded systems, and an organizer/contributor to major conferences in verification and cyber-physical systems. Notably, their group has produced formal proofs of absence of adversarial perturbations in neural networks applied inside closed-loop control—an uncommon achievement that bridges ML robustness and systems-level assurance.
11 years of coding experience
18 years of employment as a software developer
BSEE Electrical and Computer Engineering, BSEE Electrical and Computer Engineering at Rice University
PhD Electrical and Computer Engineering, PhD Electrical and Computer Engineering at University of Illinois Urbana-Champaign
French, Spanish, English