Main Navigation Menu

Design challenges of enabling an Industry 4.0 manufacturing ecosystem

IoT Design challenges

Industrial Automation and manufacturing IT design status

In a recent IHS survey of over 1000 automation and control engineering professionals of manufacturing and process industries, the following was highlighted:

  • Between 20 and 30% of a project is rework and change implementation .This work happens after detail design requirements have been signed off.
  • Asked if prototyping was used in order to capture detailed design requirements and to speed up the overall control and MES system design process, the respondents answered as follows;
    • 1/3 of respondents used manual processes (drawings, MS Word documents, Excel documents, Visio diagrams) to prototype.
    • 1/3 of respondents used custom internally created tools to prototype.
    • 1/3 of respondents used commercial off the shelf software to prototype their requirements.

Challenges  in the process automation design space

The engineering design life cycle of process automation is still largely manual.

Control system requirements capture consists of engineers writing specification documents based on input from process engineers. These documents are subsequently manually reviewed by Quality, Automation, IT, MES, Process etc. and the review cycle takes a long time. Eventually the document is signed off and an automation engineer has to interpret the pseudocode and program the chosen platform. A large number of technical queries come out of the programming effort.

Object-oriented style integration between process control systems is still not widely supported.

Internet software development and the development of secure web services , made inter-system communication much easier. As of July 2015, almost 14000 public Web API’s were registered at The author could only find 1 API from an industrial automation supplier (Honeywell).

Better data semantics alignment needed for easier communication.

In response to dissatisfaction with different proprietary systems being difficult and expensive to integrate, industrial internet consortia representing specific industry segments, are taking it upon themselves to develop the data model / data semantics for their specific industry (OPCUA companion standards) .Industry segments want greater standardisation of the data semantics needed for inter-system communication.

Greater cost certainty required.

Project managers need greater cost certainty on their software budgets as software projects have a reputation for going over budget. Is it possible to build a detailed project plan that accurately captures a complete development cycle prior to starting development?

Quicker. Cheaper. Better ; the quickening of the the manufacturing lifecycle

This demand for increased speed of manufacture and decreased time to profitability is resulting in ever more development of modelling, simulation and virtual prototyping systems. It is resulting in the increased adoption of ‘Agile’ project execution methodologies.

Insufficient automation engineering resource pool.

PLC and automation system software is expensive to design, implement and maintain,  because of the highly specific automation software skills needed to program these systems.There is a shrinking pool of automation engineers that maintain all the proprietary PLC and DCS systems possibly because companies are not investing in training the replacement engineers, possibly because its difficult to get access to the training in the first place , and possibly because maintaining old control system code architectures is not anyone’s idea of a rewarding career.

Waterfall v Agile project execution for software.

Increasingly, ‘waterfall’ software development is not seen as the best way to write software; users find it difficult to articulate requirements at the start of development because they don’t know what they want yet.Iterative lifecycles are better than waterfall for software development. Re-factoring is part of programming. Attempts to enforce waterfall execution result in increased quality assurance on what is already a poor understanding of the requirements.

Standard automation function libraries designed for re-use are platform specific.

A quality library of mature standard module templates that can be re-used time and time again, is extraordinarily valuable in terms of labour saving. This is one of the strengths of a DCS system over a PLC system. However there is no easy way currently, to import a standard template from one proprietary platform to another.(Standards such as PLCOpen do not yet have widespread support)

No automated regression testing for industrial control system software.

Amazingly, neither unit testing nor automated end-to-end testing exist yet for automation control systems.A major reason why, is that present IEC61131-3 control language implementations are simply not object oriented enough (no dependency injection implementation).

No Strong FEED (front end engineering design) without proprietary system knowledge.

A good FEED will reflect all of the client’s project-specific requirements and avoid significant changes during the execution phase. However, how can a good automation FEED take place without access to specialists trained in a particular proprietary target  automation or MES platform? Often times that platform has not been decided upon yet , or an emulation of that platform is not available. Where is the equivalent of the 3D CAD model for enabling collaboration?.

Modular process engineering implications for Automation.

Pre-planned and pre-built modules are being used more and more in place of bespoke processing engineering plant that is built in situ at the target factory. This  modular approach to plant building also requires a modular approach to automation and automation integration. How do we manage to keep control systems from different vendors aligned with a centralised standard such that they may be integrated? Namur working group NE148  was set up to answer exactly that.

Automation engineering future design trends

Unified data model.

In the factory of the future there is a unified data model  which makes it easier to integrate systems in a more  object oriented way.There is less design rework and more reuse of existing classes, with design patterns being encapsulated within  open standards to better guide the automation integration.

De-risking by modelling, simulating and testing

More and more processes are researched and developed through simulations, oftentimes with the prototype / model being tested virtually.

Industrie4.0  / enterprise IoT

Industrie (or industry) 4.0 is an IoT-like concept coined in Germany but now used throughout the world to describe the complete integration of the industrial automation component supply chain, custom manufacture & assembly, and customer requirements. IoT creates new security and integration requirements for control software.

The solution to all of these challenges starts with laying out a vision of how you want all your control and MES systems to work together; by creating a model or prototype.


Prototyping control systems

Prototyping the factory control ecosystem

Generate design specifications in record time saving money.

Better software starts with better requirements. It is impossible to demonstrate the requirements back to the client for verification before development begins, without some form of prototype.

Reduce project schedule risk by improving planning and work estimation.

Without shared understanding of a common vision, it’s difficult to plan and estimate the overall development effort.

Brain storming and idea generation

Innovative ideas are inherently uncertain and often go through several iterations to incorporate feedback before they are either realised or killed off. Teams need the ability to quickly collaborate and iterate on application prototypes to explore ideas, test hypotheses, and validate designs before even entering the development cycle.

Training benefits.

Operators can start to interact with and learn about the system before it is actually programmed.

Help expedite modular automation architecture.

Provide different suppliers with a common model that represents the target implementation. In addition, that common model can generate design specifications and test specifications in a common way for all. This also enables the ‘don’t repeat yourself’ principle and the use of corporate standards and design patterns.


The value of modelling factory floor communications stretches way beyond control and  MES system requirements capture; If the prototyping system supports Industrie 4.0 technologies such as OPCUA and AutomationML, then the communications between entire control and manufacturing IT ecosystems can be mapped out in a virtualised environment . And not just for the process industry.  HVAC systems or life safety systems can be simulated in order to build a holistic view of an ecosystem. And this simulation can then double as test server instances for real implementation integration testing through OPCUA. Or as a template for a complex rollout .Or perhaps a training rig. It can form the basis of a corporate wide benchmarking strategy.

The old carpentry adage ‘measure twice , cut once‘ says it all. The problem with process automation design lifecycle managment  has always to measure progress, against what .


(This post is an excerpt from  an extensive whitepaper on the subject of prototyping, its benefits , and its place in the modern design office. Contact HAL Software if you are interested in getting the full document.)

Comments are closed.
Optimization WordPress Plugins & Solutions by W3 EDGE