(Photo by Roland Neveu/LightRocket via Getty Images)
The success of early adopters has driven over 40% of manufacturers to implement enterprise-wide digital transformation initiatives in place by 2021, according to IDC FutureScape: Worldwide Manufacturing 2019 Prediction. Industrial IoT, Machine Learning (ML) and Applied Analytics solutions remain top of mind for global manufacturing teams given the declining contribution margins many factories face.
Finding the optimal point of entry to upgrade production processes may still present a challenge to organizations that are in the early stages of their digital transformation journey. Here are five actionable trends that you can take to the factory floor.
1. Pilots are shifting from trial to execution - Over the past few years, manufacturers have rushed to implement shiny new technologies without a clear strategy or focus. The result has been investments that don’t live up to their promise and a significant amount of unfound value in pursuing technology hype. Today, manufacturers are being much more strategic with their investments, shifting away from a trial and error approach to a focus on strategic execution. Pilot projects are being used to prove the concept and understand what it will take to scale a global rollout rapidly. The key to success is being intentional; understand what creates value for your business and build an initiative that supports those goals.
2. The growing importance of interoperability - Take a walk on the floor of most factories and you’ll see a wide range of equipment with various types and levels of connectivity. The push for smart factories has led to the next problem: interoperability. Different machines and sensors typically operate on different protocols making it difficult to centralize the data and extract insights. As a result, we are seeing more people gravitate toward the use of standard protocols such as OPC UA and MQTT. These standards translate the data into a common language allowing manufacturers to realize the benefits of a connected factory. It’s also worth noting that not every part of every line needs to be connected. Start with the business value and the problem you are looking to solve then identify which data points are needed to support the application. Just because you can collect a data point doesn’t mean you should.
3. The rise of corporate guidelines - Global manufacturers have always empowered plant managers to make decisions when it comes to equipment and vendors. However, the lack of standards has made it difficult to implement company-wide digital transformation initiatives. Over the next year, we’ll begin to see the rise of guidelines where corporate teams dictate a certain level of capability including baseline standards for connectivity, security and interoperability. Individual factories will still have the autonomy to select their vendors and partners but it will no longer be at the expense of corporate initiatives.
4. It’s no longer Cloud vs. Edge, it’s Cloud-Edge - There has been a constant debate around what is better, deploying in the cloud or on the edge. The answer is both as hybrid cloud-edge platforms grow in popularity. The cloud offers better economies of scale for storage, significantly more power for Machine Learning and Applied Analytics, and allows you to easily combine data sets from multiple sites to further advance analytical power. The edge, on the other hand, provides business continuity. Once an ML model is built in the cloud it can be deployed via the edge on the factory floor which removes connectivity dependency while increasing reliability and lowering latency. The combined solution provides the most benefits: reliability, power and cost-effectiveness.
5. Cybersecurity is top-of-mind for everyone - The biggest problem that manufacturers have is their use of legacy systems with exposed vulnerabilities. Most of the high profile ransomware attacks have used old windows security flaws as the attack vector. A common challenge is manufacturers have software that only runs on old unsupported operating systems, and they are purposefully not updating to the latest patches. This leaves glaring gaps in their security procedures. Another key attack vector is phishing, the best solution against phishing is two-factor authentication (2FA). In fact, Google removed 99% of all phishing issues through 2FA. But we rarely see manufacturers implementing simple solutions like 2FA to help address big existing cybersecurity problems.
These are actionable trends to look for in the coming year to identify cost-effective solutions to solve specific problems. With clear insights such as these, we will continue to see innovation becoming faster and waste being reduced.
