FEATURE Industry 4.0
Harness graph databases for Industry 4.0 advantages
Build a digital twin using a graph database to gain the operational advantages of Industry 4.0, advises TigerGraph’s Technical Director, Richard Henderson
T
he Fourth Industrial Revolution promises vast data exchange, backed by cyber-physical systems, the IoT, cloud
computing and machine learning. The benefi ts of Industry 4.0 include increased fl exibility and reduced manufacturing costs. However, the vast amounts of data needs to be shared and accessed by disparate systems for informed decision making, which creates challenges. As they transition to Industry 4.0, companies often build upon legacy databases which have not been designed to scale up and handle increased complexities. So then, the rapid elastic manufacturing of Industry 4.0 requires more fl exible data handling.
Smart data
Manufacturing is becoming increasingly autonomous through robotics, sensor technology and machine learning – smart machines already process materials and analyse and diagnose issues without human intervention. In addition, groups of machines now communicate, analyse tasks and work together. Enterprises are even going beyond this by creating high-level virtual representations of their businesses through ‘digital twins’, which serve as real-time digital counterparts of all or parts of the operation. Digital twins allow for a detailed analysis of the relationships between parts of the business, both internally and externally, to track inventory and assets, analyse processes,
16 September 2021 | Automation
trace faults and mitigate disruptions and system shocks. A digital twin takes inputs from across the system and gives a real-time, granular model of what’s happening, mirroring the real-life web of assets and materials. This web of relationships delivers greater operational insights and allows machines to interact more effi ciently.
Relational versus graph databases The concept of digital twins is not new, but attempts to create them in traditional relational databases quickly come up against technical limitations as the complexity of these relationships rises. Traditional relational databases require data to fi t into tables, rows and columns, typically linked to other tables through table joins. These, however, are not only diffi cult to write and debug, but are memory-intensive and progressively slower as the database expands.
Graph databases – in particular property graphs – allow an easier representation of real-world objects and processes and the running of complex algorithms. So now, rather than starting with a rigid data structure of tables, columns and rows, it begins with datapoints and links, which is more fl exible and better refl ects the real world. Hence, unlimited links to other datapoints can be created and, crucially, their relationships defi ned from any given datapoint – be it a machine or manufactured object.
Industry 4.0 graph databases Why is a graph database ideal for Industry 4.0? For a start, in I4.0, what we are modelling is usually a network of interacting processing elements which have inputs, actions and outputs. The most direct representation should resemble a network of interacting processes, and this is exactly what graph databases do. Add in real-time events, network-aware analytics and the ability to reconfi gure the network without programming eff ort, and the argument to build digital twins in graph is compelling. With a one-to-one correspondence between the real world and the database, there won’t be a need to translate data into or out of a diff erent representation, removing an entire class of work and enhancing understandability, development speed and database performance. Powerful graph analytics also allow you to traverse the database and present results in the context of the model. Tabular databases struggle to trace through networks of linked components, leaving some of the most important operational questions unanswered, such as systemic impacts of single-point failures and end-to-end constraints on resource fl ows.
Graph analytics allows you to fi nd critical paths and global and local risk metrics and perform root-cause analysis, scenario modelling and constraint optimisation, among other common operational tasks that have traditionally been diffi cult to perform, especially in real time. With a real-time graph database and graph analytics, top-level Industry 4.0 decision making can be coordinated in ways previously not feasible at scale. Furthermore, those advantages can be brought into the operational space by being effi cient, accurate and timely.
CONTACT:
TigerGraph
www.tigergraph.com
automationmagazine.co.uk
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50
