John Burtenshaw

—

Unclaimed · AI-structured profileOpenAlex

Researcher in industrial machine learning: hybrid deep models for predictive maintenance and RL-augmented PID control.

papers here

reviewed by Evolve

first-author

corresponding

—

works · OpenAlex

—

h-index · OpenAlex

Bibliometric figures come from OpenAlex (2026-10-13) and cover the author's entire public record, not just the papers listed here. Reproducibility signal comes from Evolve's own reviews.

Research profile

Output—

Not yet scored. The catalog holds too few indexed outputs to characterize publication volume. This axis would measure total papers over time, yearly output rate, and authorship-position trends, none of which current bibliometric coverage provides for this profile.

Impact—

Not yet scored. Citation and influence signals are unavailable for this profile's works. This axis would measure citation counts, h-index-style aggregates, and uptake of the two 2026 arXiv papers, which currently lack citation coverage.

Focus0/100

The focus score of 0 reflects an even 50/50 split between two adjacent areas — machine learning for industrial AI and predictive maintenance, and reinforcement learning for industrial control. The distribution spans one application cluster rather than many fields, yet the score still registers zero, so the raw split is reported as-is; both 2026 papers sit inside this hybrid industrial-ML perimeter.

Collaboration46/100

The collaboration score of 46 rests on nine distinct coauthors across the profile's cataloged work. The frequency list is dominated by Zhengyang and Gu, each recurring, alongside Thomas Cook, Joseph E. Hernandez, Sean Scott, and Chris Couch. Both 2026 papers list this researcher in a middle authorship position without corresponding status, suggesting a steady contributor within a small, recurring team rather than a solo or large-consortium pattern.

Activity—

Not yet scored. Recent-activity signals are insufficient in the current data. This axis would measure publication cadence across recent years and continuity of new work, which cannot be assessed from two same-year arXiv records alone.

Reproducibility38/100

Reproducibility averages 38 across the two papers scored. The 2026 Industrial Benchmark paper is rated low — single-episode evaluation, no significance tests, and an unproven MIMO premise — while the predictive-maintenance benchmarking study reaches medium on 700+ reported runs but is weakened by synthetic linear targets, absent significance testing, and missing linear baselines. The mean of 38 therefore reflects one low and one medium readiness rating rather than any verified artifact release.

Research summary

John Burtenshaw is a machine-learning researcher whose cataloged work concentrates on industrial applications of deep learning, split evenly between machine learning for industrial AI / predictive maintenance and reinforcement learning for industrial control. The Evolve catalog currently holds two 2026 arXiv papers, both with middle authorship positions and neither as corresponding author. The first proposes a hybrid controller that augments classical PID with TD3-discovered Gain/Shift parameters, reporting -203.46 versus -204.92 for the best deep-RL agent on the Industrial Benchmark; Evolve's review calls it promising but flags single-episode evaluation, no significance tests, and an unproven MIMO premise, with low reproduction readiness. The second benchmarks hybrid deep architectures for Industry 4.0 predictive maintenance, claiming LSTM-Transformer superiority on noise robustness across 700+ runs; evidence is weakened by synthetic linear targets, absent significance testing, and missing linear baselines, yielding a medium readiness rating. Radar dimensions give collaboration 46 (nine distinct coauthors), reproducibility 38 (mean over two papers), and focus 0 despite a 50/50 field split; output, impact, and activity axes lack sufficient data. Recurring coauthor names include Zhengyang, Gu, and Thomas Cook, alongside Joseph E. Hernandez, Sean Scott, and Chris Couch. The overall picture is an early-stage, industrially anchored profile whose claims would benefit from stronger statistical validation and released artifacts.

Collaboration pattern

Bibliographic records show nine distinct coauthors, with the frequency list repeatedly surfacing Zhengyang, Gu, and Thomas Cook, alongside Joseph E. Hernandez, Sean Scott, and Chris Couch. Both 2026 papers carry middle authorship positions without corresponding status, consistent with a technical contributor operating inside a small recurring team that straddles reinforcement-learning control and industrial predictive-maintenance work. No solo-authored output appears in the catalog.

Research keywords

deep reinforcement learningPID controlIndustrial BenchmarkTD3hybrid controlreward shapingpredictive maintenanceIndustry 4.0LSTMTransformerhybrid deep learningnoise robustness

Highlights

Hybrid PID + TD3 controller matches the best deep-RL agent on Industrial Benchmark
A 2026 arXiv study reports a PID controller augmented with TD3-discovered Gain/Shift parameters reaching -203.46 versus -204.92 for the best deep reinforcement learning agent on the Industrial Benchmark. Evolve's review calls the result promising but flags single-episode evaluation, no significance tests, and an unproven MIMO premise; reproduction readiness is rated low.
LSTM-Transformer benchmarked for noise-robust predictive maintenance in Industry 4.0
Across 700+ runs, this 2026 benchmarking study claims LSTM-Transformer hybrid architectures lead on noise robustness for Industry 4.0 predictive maintenance. Evolve notes the evidence is weakened by synthetic linear targets, absent significance tests, and missing linear baselines; reproduction readiness is medium.

Timeline

2026First cataloged output: middle-author contribution to a hybrid PID–deep reinforcement learning controller (TD3-discovered Gain/Shift) evaluated on the Industrial Benchmark.
2026Co-authored a 700+-run benchmarking study of LSTM-Transformer hybrids for noise-robust predictive maintenance in Industry 4.0.
2026Both 2026 arXiv papers place the researcher across reinforcement-learning industrial control and machine learning for predictive maintenance, within a network of nine distinct coauthors.

Papers in the Evolve catalog

Frequent co-authors

Zhengyang×2 Gu×2 Joseph E. Hernandez×2 Thomas Cook×2 Sean Scott×2 Chris Couch×2 Zhengyang×2 Gu×2 Thomas Cook×2

About this profile

This profile is AI-structured by Evolve.Pub from public bibliographic data (OpenAlex) and Evolve's own paper reviews. It is not written or endorsed by the researcher unless the profile is claimed. Radar dimensions are computed from real bibliometrics with fixed formulas — raw figures are shown alongside; missing dimensions mean insufficient data, never zero. Errors? Claim the profile to correct or hide it.

Bibliometric scores derive from OpenAlex-derived data; no citation, output-volume, or recent-activity coverage was available, so those radar axes are marked missing.
No institution, country, or ORCID is linked to this record; 'John Burtenshaw' is name-resolved from public bibliographic data and may be subject to same-name ambiguity.
This profile rests only on two 2026 arXiv papers reviewed by Evolve; any inference about productivity, impact, or career stage would exceed the evidence.
Reproducibility (mean 38) is computed from Evolve's review of two papers, not from verified code or data releases.