PRODUCT DEVELOPMENT & PROCESS ENGINEERING · SMART MANUFACTURING · GERMANY
SIDDHESH
JOSHI.
Product Development & Process Engineer
M.Sc. Advanced Manufacturing · TU Chemnitz | Smart Production Systems · Industry 4.0
Building technical systems end to end across product development, process engineering and manufacturing research - from concept CAD, DFMEA and prototyping to Python-based real-time data acquisition and ML-driven analysis.
4+ yrs
Engineering, research & development
EXPERIENCE
6
Projects across design, research & industry
PROJECTS
20 kHz
Real-time DAQ system built
RESEARCH
1
Peer-reviewed publication
PUBLISHED
01 - About
An engineer who builds
- from concept to release.
Depth to build technical systems from the ground up, and the range to deliver them as working, decision-ready solutions - engineering that adapts to the problem, not the other way around.
⚙️
Manufacturing Research & Automation
Building real-time weld process monitoring at TU Chemnitz - a Python framework that fuses high-frequency current/voltage signals (up to 20 kHz), industrial vision and thermal imaging, with ML-based anomaly detection and defect classification. Rooted in mechanical design, DFMEA and advanced manufacturing.
📋
Project Management & Technical Documentation
Driving client engineering projects end to end at Excelloit - requirements gathering, engineering calculations and technical documentation, building the project pipeline in the ERP system and digitalizing documentation workflows. Plus technical assessment of industrial machines against client and production requirements.
02 - Experience
The path so far.
Project Assistant - Automation & Process Research
Oct 2025 → Present
Technische Universität Chemnitz · Chemnitz, Germany● CURRENT
- Developing a real-time GMAW weld monitoring framework in Python, fusing high-frequency current/voltage signals with synchronized industrial-camera and thermal imaging
- Built a Python-integrated NI USB DAQ + signal-conditioning setup, scaling acquisition from 10 Hz to 20 kHz with configurable multi-rate sampling and auto-trigger
- Engineered structured, analysis-ready data logging to support ML-based autonomous welding control and real-time weld-seam analysis
- Benchmarked against NI PXIe and FlexLogger using time-aligned waveform and RMS analysis to validate accuracy and reliability
Working Student - Project Management & Technical Documentation
Feb 2026 → Present
Excelloit Consultancy Services GmbH · Mülheim-Kärlich, Germany● CURRENT
- Manage client engineering projects end to end - collecting and validating initial requirements and data, coordinating deliverables and driving each project to successful execution
- Prepare technical documentation and engineering calculations (energy, machine and process calculations, technical plans and specifications) for reliable project execution
- Build and maintain the project pipeline in the company’s ERP system and help digitalize technical-documentation workflows on a single unified platform
- Perform technical assessment of industrial machines and equipment against client and production requirements, and conduct market and competitive research to support equipment, process and investment decisions
Mechanical Design Engineer
Sep 2022 → Nov 2023
Edhaa Technologies Pvt. Ltd. · Pune, India
- Led end-to-end mechanical product development across three products - concept CAD, design reviews, DFMEA, prototyping and release documentation (drawings, BOMs)
- Owned client/supplier and stakeholder communication from concept through production handover
- Worked with the electronics team to hit critical specs - including high-precision response at 3° and 7° actuation using a magnetic-field sensing concept
- Applied DFM/DFA and supported APQP/PPAP and supplier technical evaluations
Engineering Intern
Mar 2022 → Jul 2022
International Automotive Components (IAC Group) · Pune, IndiaINTERNSHIP
- Supported production planning and manufacturing-process improvement on automotive interior components
- Contributed to line analysis that reduced cycle time and latency during complex assemblies
- Gained hands-on exposure to OEM-grade quality and manufacturing standards
03 - Research & Engineering Systems
Systems in focus.
Two systems that turn raw welding signals into real-time, actionable quality decisions - both built end-to-end in Python.
● Live acquisition GUI - synchronized V/I, camera + thermal preview, up to 20 kHz
Live Acquisition · TU Chemnitz
Intelligent Process Monitoring for Weld Defect Detection in GMAW
TU Chemnitz · Research Project · Oct 2025 → Present
Problem
Weld quality is normally verified after welding through post-process inspection - slow and costly. Defects like porosity, slag inclusion or burn-through are only caught once the part is already made.
Solution
A Python framework that interfaces multiple hardware streams in real time - NI USB DAQ with signal conditioning for current/voltage, a Baumer monochrome industrial camera and an Optris thermal camera - synchronized by timestamp. It captures the welding feature set (current, voltage, wire feed rate, gas flow, arc distance, travel speed, nozzle angle) for live analysis, with a GUI for live waveforms, configurable sampling and auto-trigger on arc voltage.
✦ Real-time multi-sensor data acquisition system
✦ Scales acquisition from 10 Hz to 20 kHz with multi-rate logging
✦ Multi-sensor fusion - electrical, vision & thermal, timestamp-aligned
✦ Architected for more sensors - foundation for ML-based autonomous welding control
↗ Read full case study
✦ Scales acquisition from 10 Hz to 20 kHz with multi-rate logging
✦ Multi-sensor fusion - electrical, vision & thermal, timestamp-aligned
✦ Architected for more sensors - foundation for ML-based autonomous welding control
● Hotelling T² & Q-residual at 99% UCL - 33 defect regions classified with confidence
ML & SPC · Industrial Project
Failure Provocation & Qualification of a Measurement Method to Realize Reproducible Arc-Welded Components
BEAS Technology GmbH · Industrial Project · Nov 2025 → May 2026
Problem
BEAS had logged weld feature data (current, voltage, timestamp) and needed an offline system to evaluate weld-seam quality from it - locating anomalies, classifying defect type, and quantifying confidence.
Solution
An independent Python GUI with a full ML workflow: load & train a PCA model, run multivariate SPC - Hotelling's T² and Q-residual at a 99% UCL - intersected with Isolation Forest for robust anomaly detection, then classify segmented defect regions with an RF / XGBoost ensemble. One-click export to PNG/PDF/SVG plots and a detailed Excel report.
✦ Flagged ~30.6% abnormal samples and segmented 33 defect regions
✦ Classifies slag inclusion, porosity, irregular penetration & burn-through with per-region confidence
✦ Locates each defect by position - built for reproducible, Industry-4.0-ready weld validation
↗ Read full case study
✦ Classifies slag inclusion, porosity, irregular penetration & burn-through with per-region confidence
✦ Locates each defect by position - built for reproducible, Industry-4.0-ready weld validation
● LCA dashboard - cost, CO² & energy across materials and manufacturing routes
Life Cycle Engineering · TU Chemnitz
Life Cycle Assessment - Stainless Steel vs. Polypropylene
TU Chemnitz · Life Cycle Engineering · Mar 2025 → Aug 2025
Problem
Choosing a material - here for a soap-holder case study - means trading off cost, environmental impact and function, decisions too often made on intuition rather than data.
Solution
Applied Life Cycle Engineering tools to compare stainless steel and polypropylene across their full manufacturing routes (blanking, stamping, injection moulding) - quantifying cost, environmental footprint and functional performance, and using ternary diagrams and sensitivity analysis to visualise the trade-offs.
✦ Polypropylene emerged as the more eco-friendly and cost-effective option
✦ Stainless steel was functionally superior but carried higher energy use and lifecycle impact
✦ Delivered a data-driven material recommendation backed by sensitivity analysis
↗ Read full case study
✦ Stainless steel was functionally superior but carried higher energy use and lifecycle impact
✦ Delivered a data-driven material recommendation backed by sensitivity analysis
04 - Hands-On Experience
From product development to project delivery.
The full hands-on cycle - three product-development projects delivered as a mechanical design engineer, plus current project-management and technical-documentation work.
05 - Skills
The toolkit.
Data, Research & ML
Python (DAQ, signal processing, automation)
Real-time DAQ (NI USB DAQ / PXIe)
Computer Vision integration
Multivariate SPC (Hotelling T², Q-residual)
PCA
Isolation Forest
Random Forest / XGBoost
Data Analysis
Excel automation
Process, Manufacturing & Quality
Process Engineering & Optimization
Manufacturing Automation
Smart Production Systems
Industry 4.0
Quality Assurance
Six Sigma (DMAIC, SPC)
Root Cause Analysis
MFD & MFA
Product Development & Design
Product Development (concept to release)
DFMEA
DFM / DFA
GD&T
Prototyping
APQP / PPAP
BOM & 2D Drawings
PTC Creo
CATIA V5
SolidWorks
AutoCAD
Project, Delivery & Documentation
Project Management & Coordination
Requirements Gathering & Management
Technical Documentation
Engineering Calculations (energy & machine)
ERP Systems
Digitalization / Digital Transformation
Cross-functional Collaboration
Stakeholder & Client Coordination
Machine Assessment
Market & Competitive Research
06 - Education & Research Output
Academic foundation.
M.Sc. Advanced Manufacturing
Technische Universität Chemnitz
Oct 2024 → Present · Chemnitz, Germany
Smart Production Systems · Process Engineering · Life Cycle Engineering · Joining Technologies & Strategies
B.E. Mechanical Engineering
Savitribai Phule Pune University
Jul 2018 → May 2022 · Pune, India
CGPA 8.38 / 10 · German equivalent: 1.8
English C1
Deutsch B1 → B2
Conference Publication
Enhancement of Automation and Robotics in a Bowling Alley for Operational Efficiency & Consumer Experience
5th Advanced Manufacturing Students Conference (AMSC) · TU Chemnitz
June 2025
Peer-reviewed conference paper proposing automation and robotics interventions to improve operational efficiency and the customer experience in bowling-alley operations.
↗ View publication (Qucosa)
Peer-Reviewed Publication
Design and Fabrication of Pedal-Powered Washing Machine
International Journal of Engineering Research and Applications (IJERA)
April 2022
CAD design, structural analysis and prototype validation of a pedal-powered washing-machine mechanism - derived from the Bachelor's final-year project.
↗ DOI: 10.9790/9622-1206016467
07 - Contact
Let's build something.
Hard problems, raw data, and the freedom to build the right solution - that's the work worth doing. If you're tackling something difficult, let's talk.
// profile.json
{
"name": "Siddhesh Joshi",
"base": "Chemnitz, Germany",
"focus": "Product Development + Process Eng.",
"education": "M.Sc. Advanced Manufacturing",
"stack": ["CAD", "Python", "NI DAQ", "ML / SPC"],
"languages": ["English C1", "German B1→B2"],
"edge": "concept to release, data to ML",
"open_to_work": true
}