Microservices in Manufacturing: Giving Operators the Tools…

Microservices in Manufacturing: Giving Operators the Tools…

Jun 7, 2016

“Microservices in Manufacturing: Giving Operators the Tools + Power they Need to Drive Efficiency”

By Andre Wegner, Founder/CEO at Authentise, LinkedIn

Full automation in production won’t be achieved in one fell swoop. Instead it’s a series of improvements that, with careful diligence and built on the right backbone, will make the long-awaited seamless art-to-part digital thread a reality. Fortunately, emergent software architecture backs this trend.

We’ve got the greatest clients in the world – they’re always looking for ways to improve and teach us a lot along the way. Here are some of the things we’ve learnt from them so far:

  1. What they have: All of them have production management systems honed over years of experience, whether it’s in Additive Manufacturing (AM) or otherwise. They have processes, often encoded in software, and experienced people to make sure things run smoothly. In short: They have proven operational excellence time and time again, otherwise they wouldn’t be in business.
  2. What they want: Automation of dreary tasks so they can focus on more interesting ones. Transparency in the effort so they can seek ways to improve (and they can represent their achievements to management).
  3. What they don’t want: Something that completely disrupts the way they work or gives them multiple system/buttons to learn and use. In short: A monolithic enterprise suite that they have to learn.

These goals and the way they go about them, put our clients in step with their organizations. By deploying automation, this unrelenting pursuit to improve makes their organizations as whole more agile. The piecemeal nature in which they deploy it also maintains the basic pillars of manufacturing: uptime and security. There is also a focus on action, not just insight, which we have also observed in manufacturing as a whole.

Nevertheless, differences abound:

  • Unlike organizations, individuals or small teams may lack access to the tools to fulfil their vision.
  • Unlike individuals, organizations don’t always know what’s needed to drive efficiency at ground zero.

Technology can help bridge the gap (finally, as one commentator pointed out), and its evolution towards microservices prove why (see box).

Microservices enable products and processes to evolve on the spot, which is appealing in a manufacturing setting that requires incremental improvement. Building a corollary to microservices in manufacturing requires us to expose sensible microservices (let’s call them modules to not confuse them with the tech term) to operators in the art-to-part spectrum who can pluck them together to address their specific needs.

A number of companies are experimenting with ways to do so. Procter & Gamble is working with GE Digital to build an app platform that allows engineers and operators to automate certain work processes. Axoom, a spin out from German machine manufacturer Trumpf, is trying to build a platform too. But apps remain monolithic entities that are hard for engineers to create by themselves. Apps don’t integrate well into existing process flows. Apps break down the unit of software, but don’t go far enough to truly give the operator more control.

Cimpress, the self-proclaimed leader in mass-customization, went a different way. Instead of offering fully blown apps – they wanted to afford their engineers the opportunity to develop apps based on their own needs so implemented a modular drag and drop interface that allowed them to create logical automation steps at the smallest possible unit. In doing so, they have taken inspiration from a number of educational tools design to lower the barrier to programming, such as:Lego Mindstorms (Lego’s cool robotics platform), Scratch (MIT’s tool to help kids develop their own videos or games with no prior coding experience) or the interface for Softbank’s robot Nao.

This is the direction that Authentise is heading in too. We’re still busy working hand in hand with our clients such as Nike and Ricoh to design and implement process automation systems for their AM workflows based (our reason for starting with AM is detailed here). But all of these are built on the 3Diax platform of modules.

We are currently working with select partners to start deploying these modules outside of just AM. Once we have a sufficiently tested set of modules and connectors (built on emerging IIoT standards such as MTConnect), and the interface to manage them, we will release it more openly. Until then, please get in touch if you are interested to join in the pilot phase.

The ultimate goal is to reduce the art-to-part timeframe to near-zero. But doing so has unintended benefits. When engineers can create their own automation tools, we can:

  • identify trends that can lead to better capital or research expenditure,
  • connect operators with similar challenges,
  • and, in the medium term, bring those automation fragments together into a continuous thread.

To do so, it’s not simply enough to architect our software with microservices, we have to expose those units to the operators who know best how to assemble them.


Links from text box:

  1. https://en.wikipedia.org/wiki/Conway%27s_law
  2. http://nigelbevan.com/papers/usweb98.pdf
  3. http://martinfowler.com/articles/microservices.html


Leave a Reply

Your email address will not be published. Required fields are marked *