Fledge is a  leading open source industrial edge platform for industrial data pipelines, intelligent edge applications and ML.

Our Vision: Fledge’s project and community deliver data management for automation, advanced process control and IIoT, accelerating global smart manufacturing / Industry 4.0. As manufacturing becomes more data driven, the convergence of OT’s & IT’s people, processes and technology gains greater importance. Like Apache Spark’s pipelines for data lakes, Fledge’s pipelines for the automation pyramid and IIoT unify data acquisition, raw integration of data, filter/clean/transform data, aggregate data and enable edge 4.0 application development, provisioning, execution and management.

 

 

Fledge data pipelines run from sensors and actuators to on-prem systems to clouds and back. Fledge applications run on bare metal, VMs, containers or K3. Fledge can execute in distributed hardware as small as Raspberry Pi 0 to a well provisioned container in the cloud.

 

 

Manufacturing data systems and processes have unique requirements when compared to hyperscale data engineering and data science tools. Fledge is architected, built and system tested to meet industrial data management demands. Fledge Architecture.

“The actionable insights that Fledge’s Edge Machine Learning has delivered to our centrifuge operations team are a game changer…”
—Jennifer Dominick, Director, Focused Improvement and Analytics, ADM

Industrial Data Pipeline Requirements

 

Fledge is not a simple Industry 3.0 data acquisition and transformation gateway. Industrial software products that connect various industrial protocols and translate them into OPC-UA or a cloud or a time series database are problematic. These legacy techniques create data silos, vendor locks, protocol locks, don’t support edge applications and can not host ML/AI on the edge.

 

 

“Fledge solves diversity and complexity issues when collecting and processing data from the process control network and beyond. OSIsoft recommends that all our industrial customers and partners begin their data pipeline and IIoT journey by integrating Fledge into their industrial 4.0 infrastructure.”
—Richard Beeson, CTO, AVEVA/OSIsoft

Fledge’s microservice-based pipelines collect from any/multiple sources, process all data types and integrate to any/multiple destinations — including MQTT, KAFKA and clouds — natively. 

Fledge resolves and democratizes Industrial data pipeline challenges including:

    • Platform Agnostic 
      • Diversity of source and destination protocols
      • Namespace mapping
      • Multiple data sources/asset of different types (PLC has time series data, accelerometer has array data requiring edge based signal processing) 
      • Diversity of networks that may also be unreliable
      • Diversity of data types (timeseries, array, image, radiometric, transactional, etc)
      • Support existing equipment and analytics systems
      • Model and application diversity 
        • Sensing, Analysis and Control Loops from manual to semi-autonomous to fully autonomous 
        • Optimization model techniques include physics and chemistry based, predictive, simulation and AI
      • Ease-of-Use and Management
        • Open extensible UI
        • No code, low code and coded application development
        • Manage distributed data paths
        • Audit changes
      • Security and Orchestration
        • Pipelines and applications behind the DMZ with no outside access
        • Adapt to plant and protocol security capabilities
        • Encryption
        • Multiple user roles
        • Optional certificate based authentication
        • Kerberos
        • EVE and K3
      • Location and Distributed compute
        • Latency, data volumes, security, compliance and cost 
        • Real time information
        • Machine edge transactions, events and alerts
        • Model Inference on the edge
        • Edge MLOps
        • Set point controls
“Fledge’s ability to collect, process, transform and integrate machine data as well as run TensorFlow Lite on the edge makes it an excellent complement to Google’s AI platform… Google is proud to contribute to the Fledge project, empowering next generation industrial processes and intelligent automation.”
—Craig Wiley, Director, Google Cloud AI

Why Choose Fledge?

      • Fledge was developed by industrial data experts for industrial use cases.  Fledge is not a general purpose edge platform used in retail and consumer applications.   Our community of industrial software and hardware experts are focused on industry 4.0 application development and industrial data pipeline challenges.  Join us.
      • Some of Fledge’s unique industrial features include:
FEATURES  BENEFITS
Microservices with Plugins
  • Highly modular reusable code
  • Faster more focused development
  • Less component test
  • No-code/Low-code development
  • Scalable, manageable
High Performance
  • Vibration signal processing 4 channels 48 kHz/channel 
  • 120K total tags
  • 120K events/sec
  • Computer vision
  • Real time edge inference with set point control
Push or Pull updates

Relay data paths 

Redundant data paths

  • Secure updates and rollbacks
  • Secure data paths
  • High availability
Open Extensible UI & Open RestAPIs 
  • Faster time to market
  • Easy to use and configure
  • No-code app development
Abstracts: assets, data sources, compute, integrations and ML models
  • Data in operational context
  • Enterprise level management and security
  • Edge MLOps
  • Intelligent automation

 

  • Open, flexible, extensible — Independent of automation equipment, systems, clouds, ML tools.  No vendor lock.  No protocol lock. Works with new and old equipment alike. 
  • Focus on your unique requirements and value add.  Fledge manages the data pipelines from sensors to clouds and back enabling 4.0 applications and multi-protocol gateway services. 
  • Cost, risk and time to market.   What open source is all about. 
  • Plants, factories and mines don’t exist in clouds – Designed to scale, secure and manage operational technologies’ distributed computing reality. 

 

“Fledge’s ability to collect, process, transform and integrate machine data as well as run TensorFlow Lite on the edge makes it an excellent complement to Google’s AI platform… Google is proud to contribute to the Fledge project, empowering next generation industrial processes and intelligent automation.”
—Craig Wiley, Director, Google Cloud AI

Join our fast growing community of industrial users, contributors, technology providers, connectors, standards groups and system integrators.

Industrial Pipelines

Public Users and Contributors

Cloud / Data Management / Edge MLOps

 

“As power systems transform to deliver energy transition successfully, grid operators must be able to monitor and interact with a variety of systems where a high-volume of data is exchanged with various protocols at high-velocity of change. These exchanges require real-time, efficient, robust, and secure communications. Fledge is the core platform of FledgePower which aims to solve the problem of multiple protocol translation for power systems operations.”
—Akli Rahmoun, Telecontrol systems expert, RTE

The fourth industrial revolution requires a new open data pipeline infrastructure integrated into the automation pyramid and IIoT. Fledge is the Linux Foundation’s solution.

 

We encourage industrial companies, software developers, integrators and suppliers to read CESMII’s (Clean Energy Smart Manufacturing Innovation Institute) Principles of Smart Manufacturing article for an overview of the market and its requirements. We concur with CESMII and Fledge embodies those principles.

 

How to cite this article: C. Leiva, J. Adv. Manuf. Process. 2022, 4(4), e10123. https://doi.org/10.1002/amp2.10123

User & Deployment Studies

UC Davis and Opus One Using Fledge to Produce World Class Wine in Safer Conditions
Download Now