5 steps to a successful rollout, part 5: Integration

This is the final part of our series on how to escape pilot purgatory and successfully scale digitization projects. You can find the introduction here, and the previous article here.

INTEGRATING DIGITAL SOLUTIONS

Concluding our series on successfully rolling out digitization projects, we look at integration. Scaling digital-industrial solutions calls for both big-picture thinking and meticulous attention to detail. Our diligence in earlier steps has set the ground work for rollout, by choosing a good use case, working with an innovation strategy, involving stakeholders, and finding added value. By all measures, our pilot project should be a success.

There's just one problem: a pilot project happens in a fishbowl. If we want our solution to swim in the deep blue sea, it's going to have to get along with a whole ocean of other fish.

We don't really expect integration at the pilot-project stage. Doing so would be unrealistic and freeze innovation—trying new things means stepping outside of what's established. But when we scale a solution, we want to integrate it with both existing systems and processes.

Integration has many benefits. In short, it reduces complexity. Consolidating and rationalizing systems and processes streamlines operations and delivers other benefits—a key topic at McKinsey's Technology IT Conference earlier this year: integration allows for quicker adaptation to change, lower risk, and reduced mid-to-long-term costs.

But these benefits come with short-term costs: integration throws up obstacles. In the worst, most immediate cases, fundamental incompatibility with existing systems could preclude using a solution at scale. Systems integration can also present simpler challenges like access issues—can users sign in with the same company login? Seemingly minor nuisances can limit uptake and alienate stakeholders, which in turn manifests as limited process integration. If processes don't adapt to a solution, we won't realize the full benefit of the solution.

Integrating a new solution is the act of joining the integrated—the complex web of personal and mechanical interconnections that exist in every enterprise. A good solution should make waves in that structure; it's up to us to mind the ripple effects.

THE ETERNAL OPERATING SYSTEM

In October 2001, Microsoft launched Windows XP. The operating system provided a familiar Windows home experience, built on top of the kernel from their venerable NT business operating systems. It was stable, fast, user friendly, and suitable for home and business use alike. It was extremely successful.

Windows XP was so popular that Microsoft was almost a victim of its own success. Even in 2009—two years after the launch of Window's XP's successor, Vista—Windows XP was still overwhelmingly the most popular version of Windows. Tired of supporting a freeloading OS, Microsoft ended mainstream support for XP that year. They ended security support in 2014, making Windows XP computers potential security risks. The message was clear: upgrade your operating system!

That should have been the end of XP. But it wasn't.

XP wouldn't die. Five years later, in 2019, nearly one in three businesses still used XP somewhere in their operations. Even now, 21 years after XP's launch, it's still here.

Why?

The answer, of course, is integration.

The particular integration issues keeping XP in play are manifold. Old hardware can still function without being up to running a resource-hungry OS. Old versions of software might preclude updating the OS without also updating the software at significant expense. In many cases, there isn't a newer version of key software that is compatible with newer operating systems. Many businesses invested in 32-bit tools that don't work with 64-bit operating systems. And whether a solution can be upgraded or not sometimes depends on whether the company that made it still exists.

Given these issues, it's understandable why so many businesses still use XP. At the end of the day, it works. And most present applications aren't even security risks—it's entirely possible to run, for example, a machine from a Windows XP computer that isn't networked.

Windows XP in 2022 provides an example of the reality of integration: it's complicated. It's at least as challenging as it is worthwhile.  Integrating our solutions into existing systems will take work, a deep understanding of those systems, and openness to available options.

SYSTEMS INTEGRATION AT ROLLOUT

In the abstract, we want our solution fully integrated. We want every point connected, with all data available on all systems, in an accessible format. In the biggest picture, this is key to digitization and automation. This is how we realize the full benefit of scaling our solution.

But reality is never so simple. Integrating systems is a big task. Launching our pilot project within an innovation strategy, we can hope for the support we need from a dedicated integration team. Barring that, there are businesses specialized in integrating systems—they can tie things together as they are, but this isn't the same as launching and sustaining an internal integration project, capable of maintaining system-wide integration as new solutions enter circulation.

Integrating a solution is like placing a link in a chain: the new link has to fit its adjacent links. If it doesn't fit, we can replace the adjacent links. But will they fit the next links along the line? And this is a multi-dimensional chain—the interconnections are more numerous and complex the further we move away from our original integration. And everything is in motion! Nobody wants to interrupt normal operations to pick apart a working system.

And this only scratches the surface of international integration, where the same data and categories are obscured by different languages, and a single business entity can siloize data behind multiple legal names.

It's good that back in step one we chose our use case based on hard data. Faced with the hard work of integration, any bright-eyed enthusiasm for this project dims. Full integration is a distant goal, guarded by real obstacles.

But those Windows XP computers running off network show a way forward: use the solution; integrate as we go.

At AiSight, we regularly see this when implementing our predictive maintenance solution: integration with SAP is desirable, but not realistic in the short term. The timeline required for SAP integration pushes the benefits out into the mid-to-long term, while the costs are real and present. So, while full integration is difficult and expensive, added value is hazy and abstract.

That's why we provide a complete solution. The AiSight predictive maintenance solution runs parallel to existing systems, with minimal integration required. We can, therefore, deliver insights—and value—on day one. This information clarifies the value of integration. With that in hand, we can approach integration as a realistic long-term goal.

In the meantime, we realize the benefits of unlimited machine uptime. That only requires integrating our solution into existing processes.

PROCESS INTEGRATION AT ROLLOUT

Each enterprise has a way of doing things, which carefully accounts for the way things are. Changing any part of the way things are should provoke change in the way things are done—this is process integration Unlike systems integration, this is not a matter of compatibility—although it can be—but a matter of potential. The demands of process integration won't halt rollout, but they can limit its potential and obscure a solution's value contribution.

Process integration presents the human side of integration. We need teams to use the solution, and to adapt their plans to maximize its benefit. This is where stakeholder engagement, again, becomes critical. People accustomed to existing processes may need training in altered processes. They'll also need to see the value the new solution adds to their processes.

Rolling out our predictive maintenance solution calls for integration into both maintenance and production processes. Maintenance processes can immediately take advantage of our solution by making better-informed maintenance plans. With the solution integrated into maintenance processes, production can plan for reduced downtime and increased overall production, thus integrating the solution into their processes.

HOW TO MANAGE INTEGRATING A DIGITAL SOLUTION

Like integration itself, managing integration is an act of simplifying complexity. The earlier steps of this series have prepared us. Now, we can methodically work through integration.

First, we want to take a step-by-step approach to integration. Integration is a big task; jumping into the middle will not help. We should establish our steps in a rollout plan. This is a great chance to think strategically, looking for the big-picture impacts of rollout, finding out how far integration should go.

Second, we need to prepare. This includes technical preparation, communication, and training. We need our stakeholders on board—IT, production teams, maintenance teams, and management.

Third, we should test and refine our rollout plan. Remember: rollout aims to replicate the success we found in the pilot project. It should, therefore, be a replicable process that we can use more than once, making each subsequent rollout easier.

Finally, we ensure that everyone is prepared for the benefits of rollout. We've seen the added value, we know what it's going to bring, and we can adapt processes accordingly.

CONCLUDING A SUCCESSFUL ROLLOUT

Here's the truth about rollout success: it's never concluded. If something works, why stop?

With a good use case, a well-supported innovation strategy, involved stakeholders, demonstrable value addition, and the ability to navigate integration, we have the tools to successfully scale a digitization project.

Why not get started on one now? Schedule a demo here.

© 2022 AiSight GmbH. All rights reserved.

Phone: +49 (0) 30 40363399

Website: https://aisight.de