Duisburg Bridge Components
Use the pattern as context only; feasibility still depends on the actual part, material, service conditions, inspection needs, and expert review.
Industrial AI / decision systems Evidence-aware by design
I build AI systems for industrial decisions that need evidence, not just predictions.
Again and again, I found that a prediction alone did not settle the industrial question. My strongest public work today is in Laser Metal Deposition and Directed Energy Deposition at Exafuse, where I connect signals, models, engineering constraints, and human judgment to the next action and the evidence needed to check it.
Start with the LMD Decision Cockpit, then follow the method and current public work.
- Laser beam: Focused energy reaches the deposition zone.
- Melt pool: The local molten region is a process observation point.
- Deposited track: Material is added along the intended repair or build path.
- Process signal: Monitoring can flag a review question, not final quality.
- Inspection evidence: Inspection and expert review establish the next defensible claim.
Decision signal // live
Monitoring
- 01SenseSignals
- 02ModelAssumptions
- 03DecideEvidence plan
- 04VerifyInspection
Evidence status
A signal is not proof.
Process signals can support review. They do not replace inspection, testing, expert review, or release evidence.
- Source note
- Public context only
LMD Decision CockpitLMD Decision Brief v1.0Conservative output
Start with a rough LMD question. Leave with a brief.
Pick the situation, mark what is known, then expose missing information, risk flags, evidence needed, and an Exafuse review route. Inputs stay in this browser session only.
- ;
- ;
Example scenario: worn steel shaft near bearing seat.
Review the worked output, start your own brief here, or open the full workbench.
Repair damaged/worn part: start with LMD Repairability Quick Check. Not enough information. Missing information and risk flags remain visible in the brief.
How I work / signal to evidence
Sense → Model → Decide → Verify
To me, a prediction is only one part of the job. The next action, the missing context, and the verification path have to be clear too.
- 01 Sense Capture the process signal and its operating context.
- 02 Model Make assumptions visible.
- 03 Decide Plan the evidence needed.
- 04 Verify Use inspection evidence.
Current proof / decision routes
LMD/DED is my current proving ground
It is where signals, materials, machine vision, robotics, repair, inspection, and engineering judgment meet in decisions with real consequences.
Start here
Current LMD/DED work: choose the question you need to structure
Repair decision Evaluate a worn or damaged component. Expose material, damage, access, finishing, and inspection gaps before repair claims harden. Start repair brief RFQ preparation Turn a rough request into a usable RFQ brief. Separate known facts, critical gaps, useful context, risks, and the next review action. Structure an RFQ Monitoring evidence Interpret a process signal without overclaiming quality. Connect monitoring context to inspection evidence and keep the release boundary visible. Review a signal Process selection Compare LMD with SLM / LPBF route signals. Start with part size, local deposition, feature complexity, tolerance, and qualification needs. Compare routes
Selected work
Three LMD decision assets from the current proving ground
These specialist artifacts show how I turn vague industrial questions into structured decision-support workflows.
LMD Quality Evidence Ladder
- Problem
- Monitoring data is often treated as release proof.
- Helps decide
- Which evidence level is needed for a claim.
- Cannot prove
- Final part quality without inspection/testing.
LMD Repairability Index
- Problem
- Repair requests arrive before material, damage, access, and inspection facts are clear.
- Helps decide
- Whether a repair candidate is worth expert review.
- Cannot prove
- Repair feasibility, release, or service safety.
LMD RFQ Toolkit
- Problem
- Vague RFQs hide missing facts and risk flags.
- Helps decide
- How to structure known facts, gaps, assumptions, and next questions.
- Cannot prove
- Replace Exafuse commercial review or engineering approval.
Documented examples
Current LMD/DED proof, with public source context
These public examples show why evidence planning matters. Commercial case details and RFQs remain with Exafuse.
Forging Hammer Repair
Use the pattern as context only; feasibility still depends on the actual part, material, service conditions, inspection needs, and expert review.
Latest field notes
Notes from the current proving ground
Short technical explanations of the assumptions, evidence gaps, and review questions that often change an LMD/DED decision.
A Prediction Is Not Yet an Industrial Decision
Why a model output needs context, a clear next action, and a verification path before it becomes useful industrial decision support.
Height Is Not Cosmetic
Bead height, build-up height, and machining allowance affect tolerance, inspection, and repair planning.
A Camera Is Not a Certificate
A camera can show a process event, but inspection and acceptance evidence decide what the event means.
Personal platform
Explore the method, current work, and research questions.
I am studying how this approach can help with decisions based on incomplete physical signals, operational risk, and human responsibility. That is a research direction, not a cross-industry deployment claim.
Commercial boundary
For commercial LMD/DED services and RFQs, use Exafuse.
My current applied LMD/DED work is carried out through Exafuse. This site shares public methods and notes; company services, case studies, and engineering review belong there.