Challenges in the journey of digital manufacturing

Digital manufacturing is forecasted to bring out massive, disruptive changes to the industrial world. Digital manufacturing is not about technology alone – it involves a complex braid of technology and business strategies across people, machines, and processes from the different departments within the organization. It is expected that digital manufacturing will result in a paradigm shift in the way that industrial organizations will conduct their business, in the way they will produce their products, and in the way they deliver will their services. The core motivation for digital manufacturing is about staying competitive, boosting revenues and enhancing profits, while at the same time, reducing costs and being agile.

As it is popularly said, “A journey of a thousand miles begins with a single step”. Many organizations believe that it is about having an idea or a use case to start with. They figure out how to execute a pilot project, invest in data platforms, tools and technical resources to complete the project. The big question here is if this is the right way to do it? In a recent survey, it was determined that 75% of the digital manufacturing projects do not get rolled out and that there is a significant gap between perceived relevance, piloting and scaling up. It is seen that the organizations do not achieve what the business wanted or what the users wanted leading to hordes of wasted effort, costs, and time. In some case, organizations wait for a perfect plan/use-case that steps up the risks of having projects with broad scopes and procrastinated values.

Clearly digital and smart manufacturing is still in its infancy and there are many reasons why organizations that are mulling a make-over must deal will face significant challenges. Let us start with the most critical challenges.

Technology and business operate in silos

There is no clear concrete connection between the data, analytics, value and the business itself. Departments do not talk to each other – the technology teams and the business teams operate in silos. Often, technical teams are interested in the technical solution and the technology rather than the business value of the technology and its application. They are obsessed with the question: “how can we use this new technology?”. There is little focus on real business problems as questions like “how can we create additional value?” are overlooked. There is little balancing between implementation costs and the value-creating potential. Broadly, there is a great lack of clarity in how analytics and data will benefit business goals and performances. There is a great deal of uncertainty around the true value of data-driven software solutions. There is not enough knowledge on how important it can be to the success of the business. There is so much hype around this that more confusion has resulted.

Status Quo Practices

It is believed that status-quo practices are effective enough. It is believed that best practices in OEE provide enough value, and advanced analytics-based solutions like predictive maintenance or anomaly detection maybe unnecessary. For instance, it is a common perception that high quality preventive maintenance programs are effective enough for high uptime in a cost-effective way. Furthermore, most employees are deeply entrenched in their daily duties and other conventional practices. They resist when it comes to new processes, technologies and practices. Change management is perceived as a threat to their position and a challenge to their responsibilities. In general, organizations can become complacent in the context of their existing legacy business practices. They may refuse to leave this comfort zone and revamp their business model and value propositions.

Lack of connection to existing theoretical and engineering knowledge

Manufacturing professionals maybe deeply familiar with their plant, the machinery, the processes and much more. There is a strong perception that their common sense and gut feelings will trump data, patterns, software and AI at any given time. They do not trust that the insights built from software and data may match or outsmart their intuition and insights. Another objection to digital manufacturing practices primarily with analytics is that current analytics practices do not necessarily accommodate the deep, expansive engineering knowledge that has been built over decades of industrial practices. For instance, it is difficult to model and predict failures like infant mortality purely based on data, especially given that only 10-15% of asset failures are age-related and that 85-90% of the failures are random.

Lack of expertise

In recent surveys, it has been seen that the shortage of data scientists is one of the factors inhibiting the deployment of IIoT based solutions. That is, industrial companies are unable to recruit professionals that impedes their progress in building and deploying machine learning-based solutions. Furthermore, domain-experts at the plant may not have the software skills and the data engineering skills needed for implementing digital solutions. Finally, in many cases, organizations are unable to scale beyond their “PoC mode” for their smart manufacturing applications as they do not have access to appropriate tools or the skills needed. Adding to this is that there is no common roadmap or standard for deploying industrial analytics solutions.

These factors perpetuate a great deal of uncertainty and doubt among business leaders, managers and engineers in the organization. Clearly, digital manufacturing is risky and requires a great deal of audacity. But not undertaking this digital journey with the right tools and playing the waiting game is even riskier. Given the formidable challenges facing the industry, the consequences of not working with the right tools for digital manufacturing can be staggering for organizations.

  • The organization may lose their competitive advantage.
  • They are unable to collect statistics about their assets, processes, products and customers.
  • They will find it difficult to retain market share.
  • They will experience lower profitability given the increasing costs of industrial maintenance and production.
  • The organization will miss out on their chances to improve production efficiency and machine uptime performance.

Finally, early adopters of digital manufacturing will have a learning advantage on the competition. Through testing and learning, organizations can get better insights, and update products and policies faster. The organization will miss the opportunity to transfer the captured expert knowledge to the next generation of engineers.

So, how can the debates and concerns around digital manufacturing strategies be addressed? How can an organization undertake a continuously evolving digital manufacturing journey with reliable and scalable applications? How can an organization be lean in their digital manufacturing implementation with faster learnings and payoffs? We will write about these in another post. Stay tuned.

About The Author

Rajet Krishnan is the CTO and a co-founder at Viking Analytics. He is a leading expert in the areas of machine learning, wireless communications system, signal processing and statistical inference. He completed his MS from Kansas State University, USA in 2009 and his PhD from Chalmers University of Technology, Sweden. He is an entrepreneur and has co-founded successful product start-ups and an investment company. He lives in Sweden with his wife and two beautiful boys.