Conundrum Industrial AI Industrial automation based on auto ML and GANs to address predictive maintenance, quality control and optimization
Agenda 1. About Conundrum 2. Solution overview 3. Cases 4. Technology deep dive 5. Conclusion
About Conundrum Conundrum Industrial Limited is an international technology company focusing on research in AI and development of its proprietary machine learning technologies and software products for industries. Backed by Speedinvest, Austrian based VC. Overview 17 completed projects 3 deployments Wins in industrial competitions: Schneider, Aramco, Gazprom Neft, CERN
About Conundrum Conundrum Industrial Limited is an international technology company focusing on research in AI and development of its proprietary machine learning technologies and software products for industries. Backed by Speedinvest, Austrian based VC. Overview Industries Metal & Mining 17 completed projects Pipes manufacturing 3 deployments Chemical Wins in industrial competitions: Schneider, Aramco, Oil & Gas Gazprom Neft, CERN Pulp & Paper
About Conundrum Conundrum Industrial Limited is an international technology company focusing on research in AI and development of its proprietary machine learning technologies and software products for industries. Backed by Speedinvest, Austrian based VC. Overview Industries Our partners Metal & Mining 17 completed projects Pipes manufacturing 3 deployments - our investor Chemical Wins in industrial competitions: Schneider, Aramco, Oil & Gas Gazprom Neft, CERN Pulp & Paper
Improving Industries Performance Reduce maintenance costs & Predictive & prescriptive improve throughput maintenance Manufacturing process Reduce manufacturing costs optimization Eliminate quality escapes Quality control
7
Conundrum System Provides the Following Solutions: • Predictive & Prescriptive maintenance • Quality control • Industrial processes optimization
Data Sources Equipment sensors: Temperature Pressure Vibration Acoustics Amperage etc. 9
Data Sources Equipment sensors: Measurements in laboratories: Temperature Chemical analysis Pressure Product quality tests Vibration Acoustics Amperage etc. 10
Data Sources Equipment sensors: Measurements in laboratories: Temperature Chemical analysis Pressure Product quality tests Vibration Acoustics Amperage etc. Complex information Video Ultrasound X-Ray 11
Data Sources Equipment sensors: Measurements in laboratories: Temperature Chemical analysis Pressure Product quality tests Vibration Acoustics Amperage etc. Tracking of events: Complex information Failures Video Manual commands Ultrasound Control values management X-Ray Operating modes 12
Training in Cloud & Deployment in Cloud Training Module in Cloud Factory / Asset / Company Data Center Automated Preparation of Solution Production Equipment Historical Data Automation System Data upload Industrial Protocol Conundrum Training Module Prepare model and assemble solution Conundrum Inference Module Production Ready Deployed 13
Training in Cloud & Deployment in Cloud Training Module in Cloud Factory / Asset / Company Data Center Automated Preparation of Solution Production Equipment Historical Data Automation System Data upload Industrial Protocol Data Stream Conundrum Training Module Prepare model and Conundrum Cloud assemble solution Adapter Data Stream Communicate with Conundrum Conundrum Cloud Inference Module Production Ready Deployed Inferences Engineers 14
Training in Cloud & Deployment On-Premises Training Module in Cloud Factory / Asset / Company Data Center Automated Preparation of Solution Production Equipment Historical Data Automation System Data upload Industrial Protocol Conundrum Training Module Prepare model and assemble solution Conundrum Inference Module Production Ready 15
Training in Cloud & Deployment On-Premises Training Module in Cloud Factory / Asset / Company Data Center Automated Preparation of Solution Production Equipment Historical Data Automation System Data upload Industrial Protocol Data Stream Conundrum Training Module Prepare model and Conundrum Conundrum assemble solution Integration To Platform Inference Module Production Conundrum Production Ready Supercomputer Conundrum Deployed Inference Module Production Ready Inferences Engineers 16
Conundrum Platform On-premise Training Module in Cloud Factory / Asset / Company Data Center Automated Preparation of Solution Production Equipment Historical Data Automation System Data upload Industrial Protocol Data Stream Conundrum Training Module Conundrum Prepare model and Platform Conundrum assemble solution Integration To Platform Production Conundrum Supercomputer Conundrum Inference Module Production Ready Inferences Engineers 17
Training Systems Automated Transfer Learning Machine Learning Conundrum Cloud Conundrum Cloud enables to automatically prepares a use models experience and model for specific equipment transfer knowledge from pretrained models Delivers max benefit extracted from industrial data
Inference Systems Permanent Control & Production Ready Improvement in Production Auto monitoring keeps the Conundrum Inference Module performance under control provides API to simplify integration & deployment Permanent updates keep the system up-to-date Packed into container enabling platform independence
20
17 Successful Cases Across Industries Chemical Pulp and Paper Oil and Gas Metal & Mining Industries Predictive maintenance of Quality prediction of coal in Paper tears prediction on Predictive maintenance of chemical production coal enrichment process paper mill submersible pumps in oil equipment wells (upstream) Optimization of coal Kappa number (paper Solved cases Prediction of produced enrichment process quality) prediction on paper Anomaly detection of pumps product quality production (downstream) Defects detection and Optimization of chemical classification of metal sheet Prediction and classification plant operating modes during zinc coating of drilling rigs failures events (upstream) Digital twin of chemical Predictive maintenance of production welding equipment
Case: Large-Diameter Pipes Manufacturing Challenge Manufacturing process downtime due to welding equipment (DSAW) failure. Tasks - Predict failures of welding equipment - Detect causes of potential failures - Speed up check-ups and maintenance Results Before 9% After 6% 25% 61% 72% -32% Avg. check-up Downtime Detected failures True check-ups time reduction reduction
23
Case: Coal Enrichment Optimization Overview Coal mining and processing enterprise with coal enrichment factory. It sustains quality issues of coking coal which leads to losses. Tasks - Predict coal quality (ash conc.) - Provide real-time recommendations to adjust control parameters to improve product quality Results ±0.3% 88% -40% Lower number of Conc. prediction True detected quality shipments error issues with quality issue
Predictive Maintenance of Pumps Challenge Pumps downtime and extra maintenance costs due to abrupt breakdowns. Pumps quantity is hundreds of units. Avg. breakdown events per year is 3. Tasks - Predict abnormal behaviour of pumps several days before breakdowns - Detect causes of breakdowns - Speed up check-ups and maintenance Results -31% 95% 62% Avg. maintenance Detected failures True check-ups time reduction 25
Chemical Production Optimization Challenge Fertilizer manufacturer has issues of low quality of product which is estimated as a P2O5 concentration. Tasks - Predict P2O5 conc. in the output of flotation 1 hour in advance - Recommend control parameters adjustment to keep the conc. into required range Results ±0.5% 88% 20% Conc. prediction True detected quality Quality issues error issues decrease 26
PdM of Paper Mill Challenge Failures of paper mill happen which lead to paper tears and downtime of the whole mill. Tasks - Predict failure of the mill; - Recognize possible causes of failures. Results -24% 58% 62% Avg. maintenance Detected failures True check-ups time reduction 27
28
Training Module Nvidia Docker Build/Rebuild Digital Twin Get Data Auto ML, From Storage RCGAN, Attention Method Nvidia GPUs 29
Training Module Nvidia Docker Train\Retrain additional layers Build/Rebuild Digital Twin Supervised Learning Get Data Auto ML, From Storage RCGAN, Attention Method Nvidia GPUs 30
Training Module Nvidia Docker Assemble and Deploy the Model Docker Optimize Thresholds Check Quality Recognize Causes Inference Control Policy Train\Retrain additional layers Optimization, Clustering Build/Rebuild Digital Twin Supervised Learning Get Data Auto ML, From Storage RCGAN, Attention Method Nvidia GPUs 31
Technology Workflow Model Data MetaData Task Definition Uploading Data Architecture Preprocessing Generation Generation 32
Technology Workflow Model Data MetaData Task Definition Uploading Data Architecture Preprocessing Generation Generation Train Digital RCGAN Twin 33
Recommend
More recommend
