dennisthiessenandClaude Opus 5
239129b37e
Explore SDU Center for Energy Informatics PhD, Theme 2 (Predictive Maintenance and Asset Management of Smart Energy Networks), Odense DK, deadline 2026-08-20. JD scraped live via Playwright. Verified the master's thesis against the source PDF rather than the stored title alone: - Vibration-based condition monitoring of CNC machine tools; features are the dimensionless waveform, peak, pulse, margin and kurtosis indices. - Implemented a hybrid of rule-based reasoning and a 7-10-3 ANN. - Evaluation measured throughput and latency, not model accuracy: ~500 SPS pipeline against 72.9 kSPS sensors. - Corrects an earlier note in the same session: PSO was surveyed only and never implemented. Recorded under thesis_limits alongside the absence of real operational data and accuracy figures. Also records the ECTS relative grade (B, top 35%), the existence of English-language transcripts, and the concrete Bosch sensor-data types (defect management records, wafer inspection images, PCM electrical parameters) under BS-2. Adds the enquiry email sent to Prof. Bo Norregaard Jorgensen, which asks whether an industry candidate is viable and whether non-academic letters of recommendation are accepted. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
4.1 KiB
Informal enquiry — SDU Center for Energy Informatics, PhD Theme 2
To: bnj@mmmi.sdu.dk Subject: PhD Theme 2 (Predictive Maintenance) — enquiry from an industry data engineer
Deadline: 20 Aug 2026. Verified against the thesis PDF on 2026-08-02 — see
claims.json EDU-MENG for what may and may not be claimed.
Dear Professor Jørgensen,
I am writing regarding the PhD positions in Energy Informatics (job ID 4159), specifically Theme 2 on predictive maintenance and asset management of smart energy networks. I come from industry rather than academia, and I would like to ask you directly whether that is a realistic profile for these positions before I invest in the full application.
My master's thesis (M.Eng., Universität der Bundeswehr München, carried out at Tongji University Shanghai, graded 1.0) built a remote fault diagnosis system for CNC machine tools from vibration sensor data, using the dimensionless waveform, peak, pulse, margin and kurtosis indices as features. I deliberately combined two diagnosis methods: rule-based reasoning for interpretable physical thresholds, and a small feed-forward neural network for the noisy cases that static rules cannot separate — an early attempt at the interpretability trade-off that Theme 2 also raises. It was a methods prototype validated on benchmark and synthetic data, not on operational data, and its most useful finding was a negative one: the diagnosis pipeline sustained roughly 500 samples per second while the vibration sensors it was meant to serve delivered 72.9 kSPS. The model was never the bottleneck. The data path was.
I then spent thirteen years building exactly that data path. At Bosch Semiconductor Manufacturing Dresden I worked on analytics for fab sensor and process data — defect management records, wafer inspection images, and electrical parameters from process control monitoring — including integrating containerised ML inference into a 24/7 production environment. I am currently a data engineer at Swisscom, where I build and own governed data products within the company-wide data mesh on AWS, together with the pipelines and the metadata and lineage behind them. That work is explicitly the foundation layer for the AI strategy the company intends to build on top, and the argument I make daily is much the same one Theme 2 makes for energy infrastructure: trustworthy, well-described operational data is the precondition for anything downstream.
Theme 2 lists sensor measurements, operational data, inspection records and asset information as the sources to be combined. Integrating precisely those categories into something reliable enough for other people to make decisions on has been my job for over a decade — but always under delivery pressure, where the honest answer to "how well does this actually generalise" is usually "we did not have time to find out". A PhD is the deliberate move back into a setting where that question can be answered properly, and I would like to bring the operational side of the problem with me rather than leave it behind.
On location: I am fully aware the positions are in Odense and I see the move as part of the attraction rather than a cost. I have lived and worked in Norway, I hold German citizenship and currently live in Switzerland, and returning to Scandinavia is something I would genuinely welcome.
Two questions, if I may:
-
Does the centre see a candidate coming from industry after this long as a realistic fit, or does the assessment process in practice favour applicants moving directly from a master's programme?
-
The application asks for two letters of recommendation. My academic referees are thirteen years in the past. Would letters from current and former managers and colleagues be acceptable, or must at least one come from academia? I also hold German employment references (Arbeitszeugnisse), though I appreciate these are a national format that may not travel well.
I would be glad to send my CV if that is useful.
With kind regards,
Dennis Thiessen, M.Eng. Bern, Switzerland dennis@thiessen.io · +41 795 955 585 linkedin.com/in/dennis-thiessen