Jedat Inc. · TSE:3841
Jedat — Pollen Metrology — Technical Presentation (March 2026)
34 slides from Jedat's technical presentation, Mar 3, 2026
Slide 1: Webinar #4 Collaborative AI Software Layer for Process Optimization / Digital Twin
Webinar #4 Collaborative AI Software Layer for Process Optimization / Digital Twin POLLEN WEBINAR #4 - COLLABORATIVE AI FOR PROCESS 3/3/2026 1 OPTIMIZATION
Slide 2: Speakers
Speakers POLLEN WEBINAR #4 - COLLABORATIVE AI FOR PROCESS 3/3/2026 2 OPTIMIZATION
Slide 3: Webinar #4 – Agenda
Webinar #4 – Agenda Duration: 45min Collaborative AI Software Layer for Process Optimization 30min Introduction to Collaborative AI used in industrial environments Overview of how state of the art AI and Data Science techniques can improve process development and understanding while boosting process
Slide 4: 1. Context of Data Analytics for Semiconductor Process Optimization
1. Context of Data Analytics for Semiconductor Process Optimization POLLEN WEBINAR #4 - COLLABORATIVE AI FOR PROCESS 3/3/2026 4 OPTIMIZATION
Slide 5: Semiconductor Ecosystem Data Flow
Semiconductor Ecosystem Data Flow Traditional Engineer-Driven Materials Development From fragmented data to actionable process intelligence Without shared analytics, engineers cannot fully leverage either wheel Simulation R&D and Production line Process Process Process Process parameters modelling t
Slide 6: Semiconductor Ecosystem Data Flow
Semiconductor Ecosystem Data Flow Traditional Engineer-Driven Materials Development From fragmented data to actionable process intelligence Without shared analytics, engineers cannot fully leverage either wheel Exploding Data Complexity Multiple inspection and metrology tools generating heterogeneou
Slide 7: Semiconductor Ecosystem Data Flow
Semiconductor Ecosystem Data Flow Traditional Engineer-Driven Materials Development From fragmented data to actionable process intelligence Without shared analytics, engineers cannot fully leverage either wheel Exploding Data Complexity Data Silos Across the Chain Lithography, etch, deposition, bond
Slide 8: Semiconductor Ecosystem Data Flow
Semiconductor Ecosystem Data Flow Traditional Engineer-Driven Materials Development From fragmented data to actionable process intelligence Without shared analytics, engineers cannot fully leverage either wheel Exploding Data Complexity Data Silos Across the Chain AI Remains Experimental Many pilots
Slide 9: Semiconductor Ecosystem Data Flow
Semiconductor Ecosystem Data Flow Traditional Engineer-Driven Materials Development From fragmented data to actionable process intelligence Without shared analytics, engineers cannot fully leverage either wheel Exploding Data Complexity Data Silos Across the Chain AI Remains Experimental Engineering
Slide 10: Semiconductor Ecosystem Data Flow
Semiconductor Ecosystem Data Flow Traditional Engineer-Driven Materials Development From fragmented data to actionable process intelligence Without shared analytics, engineers cannot fully leverage either wheel Exploding Data Complexity Data Silos Across the Chain AI Remains Experimental Engineering
Slide 11: It needs a Transversal Open Collaborative AI layer
The industry does not need another AI tool It needs a Transversal Open Collaborative AI layer POLLEN WEBINAR #4 - COLLABORATIVE AI FOR PROCESS 3/3/2026 11 OPTIMIZATION
Slide 12: 2. Our vision and mission
2. Our vision and mission POLLEN WEBINAR #4 - COLLABORATIVE AI FOR PROCESS 3/3/2026 12 OPTIMIZATION
Slide 13: Pollen - Collaborative AI for Process Control
Pollen - Collaborative AI for Process Control AI-Driven Software for Next-Generation Manufacturing Bridging Data Silos in Material Science with Collaborative AI for Yield Enhancement • Software Company, founded in 2014, • Location: Grenoble – France • Staff Members: 20 (Data Science, Software, Proce
Slide 14: A Transversal Collaborative AI Layer Across the Semiconductor Value Chain
A Transversal Collaborative AI Layer Across the Semiconductor Value Chain Bridging Metrology, Inspection and Process Intelligence POLLEN WEBINAR #4 - COLLABORATIVE AI FOR PROCESS 3/3/2026 14 OPTIMIZATION
Slide 15: Augmenting Engineers With a Continuous AI-Driven Development Loop
Augmenting Engineers With a Continuous AI-Driven Development Loop But maturity does not come from AI alone — it comes from how AI is integrated into engineering loops Simulation R&D and Production line Process Process Experts Process Process parameters modelling DOE tool logfile knowledge Wheel of f
Slide 16: Example 1 - Smart3 Integration for Fabs
Example 1 - Smart3 Integration for Fabs Use the the SDK & the API POLLEN WEBINAR #4 - COLLABORATIVE AI FOR PROCESS 3/3/2026 16 OPTIMIZATION
Slide 17: Example 2 - Smart3 Integration for Tool Suppliers
Example 2 - Smart3 Integration for Tool Suppliers Use the the SDK & the API Production Tool supplier R&D /dev Use cases Model training Algorithm development Tool supplier software platform In- house software Inference Testing Inference CI/CD (model /algorithm Analyzed Data (model /algorithm executio
Slide 18: Process Optimization & Digital Twin
Let’s now apply it to Process Optimization & Digital Twin POLLEN WEBINAR #4 - COLLABORATIVE AI FOR PROCESS 3/3/2026 18 OPTIMIZATION
Slide 19: 3. Industry challenges and AI for Defectivity
3. Industry challenges and AI for Defectivity POLLEN WEBINAR #4 - COLLABORATIVE AI FOR PROCESS 3/3/2026 19 OPTIMIZATION
Slide 20: Problems to solve
Problems to solve Focused on Process Optimization Time-pressure - Long inspection time - Costly expert time (engineers, data scientists) Process complexity - Processes are involving a lot of different parameters - Simulation limits are reacht for advanced processes Low ROI from data - Data underused
Slide 21: Problems to solve
Problems to solve Focused on Process Optimization Time-pressure Data Cleaning - Long inspection time - Costly expert time (engineers, data scientists) Process complexity - Processes are involving a lot of different parameters Modeling - Simulation limits are reacht for advanced processes Low ROI fro
Slide 22: Daily Challenge
Daily Challenge Process Optimization is a key work of process engineer from R&D to Production. Challenge: - User have a process with input parameters (time, temperature, pressure, gas concentration, gas type, …) - The process produces a material or a device with characteristics that must be optimize
Slide 23: Data Cleaning
Data Cleaning Clean data is key to create good models. Examples of bad data: • Correlated inputs • Missing values • Constant or repeated inputs Gains: Data cleaning is the preparation to obtain good data. • Selected most relevant features • Explain dependencies • Visualize data • Prepare data for mo
Slide 24: Modeling
Modeling Modeling is the step to digitally reproduce the process. Data-based modeling is time computation efficient (compared to simulation) Can account for any number of inputs and outputs Challenges: • Data variability • Complex interaction • Low number of experiments Random Forest Gains: • Generi
Slide 25: Modeling and Physics-based simulation
Modeling and Physics-based simulation Physics-based simulation can be integrated in the global pipeline for improved accuracy. Machine Integration can be done at various levels depending on the Inputs AI Model models, simulations and available physical equations. Measured Predicted Outputs Outputs C
Slide 26: Optimization
Optimization Optimization is the step to efficiently search with the model the best inputs according user specifications. Optimization consists in running predictions in a way to find optimal points of the model depending on user specifications. Challenges: • Complex processes • Highly non convex sp
Slide 27: Design Of Experiments
Design Of Experiments Efficient Design Of Experiments allows to obtain more knowledge with fewer experiments. Data-based modeling is time computation efficient (compared to simulation) Challenges: • Experiments are costly • Large parameter space to explore • Need to maintain understanding of the pro
Slide 28: 4. Use Cases & Demo
4. Use Cases & Demo POLLEN WEBINAR #4 - COLLABORATIVE AI FOR PROCESS 3/3/2026 28 OPTIMIZATION
Slide 29: Demo plan
Demo plan 1. Collaborative Platform ➔ Example Python model 2. Data Cleaning pipeline 3. Modeling 4. Optimization POLLEN WEBINAR #4 - COLLABORATIVE AI FOR PROCESS 3/3/2026 29 OPTIMIZATION
Slide 30: Demo Video
Demo Video SmartYield3 Full presentation of SmartYield3 pipeline with the import of Custom Modeling and Optimizer https://www.youtube.com/watch?v=4R1Hd8g7CIs POLLEN WEBINAR #4 - COLLABORATIVE AI FOR PROCESS 3/3/2026 30 OPTIMIZATION
Slide 31: 5. Next Webinars, Q&A
5. Next Webinars, Q&A POLLEN WEBINAR #4 - COLLABORATIVE AI FOR PROCESS 3/3/2026 31 OPTIMIZATION
Slide 32: Next Webinars
Next Webinars Webinar 1 – Dec 2 (2025): Collaborative AI for Next-Gen Semi Manufacturing - A Software Platform Revolution Webinar 2 – Feb 3 (2026): Collaborative AI Software Platform for Metrology Webinar 3: Monday Feb. 16: Collaborative AI for Automatic Defect Classification Webinar 4: Tuesday Mar.
Slide 33: Any question?
Any question? Seminar slides access , please send an email to contact@pollen-metrology.com POLLEN WEBINAR #4 - COLLABORATIVE AI FOR PROCESS 3/3/2026 33 OPTIMIZATION
Slide 34: Thank you
Thank you POLLEN WEBINAR #4 - COLLABORATIVE AI FOR PROCESS 3/3/2026 34 OPTIMIZATION
Pollen Metrology — Technical Presentation (March 2026) on 6ix