A world-leading manufacturer was in search of an IoT solution that could ingest, integrate, and manage data being generated from various types of connected machinery located on factory floors around the globe. The company needed to manage the devices generating the data, integrate the flow of data into existing back-end systems, run advanced analytics on that data, and then deliver services to generate real-time decision making at the edge.
In this session, learn how Clairvoyant, a leading systems integrator and Red Hat partner, was able to accelerate digital transformation for their customer using Internet of Things (IoT) and machine learning in a hybrid cloud environment. Specifically, Clairvoyant and Eurotech will discuss:
• The approach taken to optimize manufacturing processes to cut costs, minimize downtime, and increase efficiency.
• How a data processing pipeline for IoT data was built using an open, end-to-end architecture from Cloudera, Eurotech, and Red Hat.
• How analytics and machine learning inferencing powered at the IoT edge will allow predictions to be made and decisions to be executed in real time.
• The flexible and hybrid cloud environment designed to provide the key foundational elements to quickly and securely roll out IoT use cases.
Strategies for Unlocking Knowledge Management in Microsoft 365 in the Copilot...
Delivering digital transformation and business impact with io t, machine learning and a hybrid cloud
1. Delivering Digital Transformation and Business Impact
with IoT, Machine Learning and a Hybrid Cloud
Franco Potepan
Director, Product Management
Eurotech
Robert Sanders
Director, Big Data Engineering
Clairvoyant LLC
2. About Clairvoyant
Background Awards & Recognition
Boutique consulting firm centered on building data solutions and products
All things Web and Data Engineering, Analytics, ML and User Experience
to bring it all together
Support core Hadoop platform, data engineering pipelines and provide
administrative and devops expertise focused on Hadoop
Data
Devops
Data
Engineering
Data
Science
Product
Design
3. About Eurotech
Eurotech is a global company that designs, creates and delivers
full Internet of Things solutions, including services, software and
hardware to leading systems integrators and enterprises.
25+ years of experience in embedded, IoT, and distributed systems:
• Embedded computing & communication solutions – standard, custom, COTS
• High Performance Embedded Computer boards & sub-systems
• Low power and ultra-low power embedded computer boards and sub-systems
Behind the products of more than
20 Global 500 companies
4. Agenda
• Current Manufacturing Process
• Opportunities for Improvement in Technology and Business
• Predictive Maintenance
• Building an IIoT Infrastructure
• Hybrid Cloud
• Eurotech Stack
• Central Data Lake
5. Current State
• Machines are running on PLCs
• Production lines feeding local databases
• PostgreSQL, MySQL/Maria, Oracle, SQL Server, DB2, etc
• These databases may feed larger Data Warehouses, Operational Data Stores or Data
Marts
• Visualization and Reporting solutions often need to tap many of these databases to gather
metrics
• Generated metrics are available to the end user a while after they’re initially generated (days,
weeks, or months)
6. Current State Limitations
• Limited End-to-End Visibility with current solutions
• Manufacturing vendor tools often look at current settings and past performance
• Lack of integration between manufacturing equipment
• Predictive capabilities are rarely included
Information lags are too long and rely on humans for interpretation, often long after an
unplanned outage occurs!
8. Business Opportunities in Industry 4.0
• Improved Productivity & Output
• Enhanced Customization for Flexible Manufacturing
• Improve Quality Control
• Access to Data Across the Supply Chain for Better Decision Making
• Predictive Maintenance
• Many more…
9. The Cost of Unplanned Outages
• 82% of companies have experienced at least one unplanned downtime outage over the past 3
years (average number of outages = 2).
• Based on Aberdeen's independent research of unplanned downtime costing companies
$250K/hour, this equates to more than $2 million.
Source: GE Digital, “After The Fall: Cost, Causes and Consequences of Unplanned Downtime”
12. Predictive Maintenance with Machine Learning
• Based on Historical Data, Age of Equipment and other factors, predict whether a machine will
fail in a certain window
• The window could be a day, week or month
• Project remaining life of your equipment
• Regression problem
• Anomaly Detection
• We compare current metrics to what’s considered normal operation
19. • Supported by RedHat
• Enterprise Kubernetes
• Container Orchestration System
• Deploy Applications and
Microservices
• Easily Scale
• Extensible
• EveryWare Cloud is designed to
run on OpenShift
OpenShift
21. Cloudera Data Science Workbench
• Cloudera’s Data Science Offering
• Provides a UI to write Spark jobs
• Spark ML
• Languages
• Scala
• Python
• R
• Docker as the underlying framework
for starting up Sessions for the user
to run processes on
• Machine Learning Models as REST
Services
• Simulations
22. Data flow to derive deep business insights and actionable intelligence
IoT Machine Learning End-to-End