Driving Industry 4.0 at Distributed Edges with Cloud Orchestration
IT / OT convergence introduces design constraints to the edge . Typically , OT and IT solutions have different constraints and requirements that guide design and implementation . While OT focuses on safety , efficiency , and continuity of manufacturing processes , IT prioritizes speed , security , and analysis of data [ 13 ]. Modern edge computing must merge the qualities of both paradigms to maximize its potential . For example , how to leverage the agility of IT systems and ensure their high availability for the continuity mission critical and latency sensitive process at OT level . Additionally , challenges regarding backend data integration at the edge from different data sources ( cloud , other edge nodes , and assets ) and their semantics that are distributed within various systems on the shop-floor must be considered . These challenges raise the question on what data is business-critical and must be computed at the edge .
Integration of legacy systems ( Brownfield ). Existing equipment might not support greenfield scenarios using the newest information & communication technologies ( ICT ). Edge computing solutions must consider strategies such as retrofitting to support legacy scenarios and semantics .
Development , Integration and Operational Costs . Distributed systems without centralized management and orchestration strategies can have a high Total Cost of Ownership ( TCO ) impact . Development costs also need to be considered to enable custom development at the edge by providing the optimal balance between standardization , openness , and configurability of applications and services . Operational approaches need to be evaluated from a variety of perspectives . One perspective is the cost related to communications costs . Filtering and aggregation capabilities reducing the workload are decisive in this context .
In the next sections , we describe how we derive these challenges into requirements that drive to solutions for edge computing as elaborated in this paper .
REQUIREMENTS FOR EDGE COMPUTING SOLUTIONS IN MANUFACTURING
This section introduces requirements for functional and non-functional features of edge computing solutions addressing the challenges above .
Features fulfilling functional requirements to build edge computing capabilities are :
• Machine orchestration
• Connectivity of edge nodes to assets , other edge nodes , and cloud applications using industry standards ( e . g ., OPC-UA , MQTT , Modbus , REST , WebSockets )
• Edge analytics , data processing , and local storage
• Zero-touch device provisioning and onboarding
• Backend data integration and semantics ( master and transactional data )
• Compliance to data security , privacy , sovereignty e . g ., General Data Protection Regulation ( GPDR )
Components fulfilling non-functional requirements to build an edge computing infrastructure and platform are :
- 22 - June 2021