- Register


Home>IIot & Smart Technology>Industry 4.0>Software-centric

Editor's Pick



09 August 2023

Peter Weckesser looks at software-defined technologies and asks if they could be the next step in digital transformation

IN INDUSTRIAL settings digital transformation can prove challenging. While 90% of industrial companies believe digitisation offers more opportunities than risks, estimates suggest that 84% of digital transformation projects fail. One of the key causes? Automated solutions can be inflexible and difficult to change.

In effect, changes can be complex and costly due to rigid, legacy infrastructure. Some estimate that reconfiguration downtime to accommodate a new run can cost north of $1 million an hour. In the same realm, machines that tightly couple hardware with intelligence can make modular production and product variance impossible. When reprogramming for production changes becomes a challenge, businesses may find themselves locked-in; falling behind on growth opportunities, competition, or their own sustainability objectives.

Software-defined approach

Current IoT open architectures address many digital transformation challenges by augmenting existing automation infrastructures and generating new data that is made available to applications. In reality, they often fail to satisfy the resiliency and flexibility needs imposed by current business imperatives. Cutting through these common barriers is the use of software-defined technologies.

A software-defined approach shifts the logic and intelligence traditionally embedded in the hardware itself to software. In many industries the hardware is already managed independently from the software infrastructure running it, that is equally isolated from the data management and from the applications management. The classic example is how servers have been abstracted: today compute, storage, and network capabilities are naturally discussed in full independence of one another.

So, what does a software-defined approach look like in practice?

Software-defined engineering

In a software-defined system, whether it is a complete plant, building, or a production line, its behaviour can be tested in a virtual environment thanks to portable, interoperable digital twins. They can simulate domains such as power, motion, hydraulics, or thermodynamics, but also provide insights on the interdependencies between them, so that designs are optimised holistically. A good example is Veolia Water Technologies, the world’s largest supplier of water services, which now uses cloud-based data and analytics to standardise plant design. Global teams use shared, secure information and leverage simulation tools to optimise operations. By pinpointing high-risk water loss areas and quickly testing resilient designs, Veolia has cut water losses at over 100 facilities, saving 15% in energy each year with 20% higher production efficiency.

Software-defined commissioning

With a complete representation of the automation system available in a digital environment, companies can simulate their complete operations, including what-if scenarios and anticipating how the code will behave when running the assets in the field. A software-centric environment significantly shortens the commissioning time, demands less experts on-site, and supports sustainability goals by reducing repetitive tasks. 

Software-defined operations…

Guided by software, machines can be networked, configured, and updated effectively – seamlessly connected to other machines thanks to interoperable, portable, re-usable code, allowing unprecedented system flexibility.

Software-defined technologies are also foundational to autonomous operations. As an example, electricity companies are gradually embracing these technologies to make digital substations the foundational elements transitioning conventional grids into digital grids capable of satisfying prosumer requirements.

…and flexible evolutions

Software-defined systems effectively decouple the lifecycle of the hardware from that of the functionality, so that:

- extensions, upgrades, and evolutions become on-the-fly software updates, largely decoupled from hardware changes,

- lifetime expectancy of the systems is no longer dependent on the availability of hardware spares, nor the active support of operating systems,

- no major revamps are necessary making systems more evergreen (shift from CapEx to OpEx),

- we achieve multi-vendor collaboration from both a technical and a business perspective.

In conclusion, a software-defined approach can entail different technologies like virtualisation solutions, portable control, orchestration capabilities and open APIs; if they enforce a new way to design, operate and maintain the systems, a way that requires strong IT competences, they can enable resilient and flexible operations.

Peter Weckesser is chief digital officer at Schneider Electric


Key Points

  • Changes to automated systems can be complex and costly due to rigid, legacy infrastructure
  • A software-defined approach shifts the logic and intelligence traditionally embedded in the hardware itself to software
  • In a software-defined system behaviour can be tested in a virtual environment thanks to portable, interoperable digital twins