Select Page

Today there is a surge of investments in Industrial IoT startups. At the other end of the spectrum, industry behemoths such as GE and Siemens are betting their futures on the same opportunity. The question is whether we are in the middle of another tech bubble or whether digitalization will fundamentally change manufacturing as we know it.

Every survey of C-Level executives confirms the obvious: the Smart Factory is inevitable.  Big data machine learning is not a fad.  The factories that operationalize sensor data that is already collected will likely see the benefit of lower machine maintenance costs and reduced factory downtimes.

At the same time, there is much anecdotal evidence that the Smart Factory concept is stuck in the planning phase. Today, we see a disconnect between executive’s strategic vision of the Smart Factory and the reality at the facility floor level.  We refer to the Smart Factory as a critical component of the Fourth Industrial Revolution, yet much of the change can be more accurately described as evolutionary. In some cases, the change is non- existent. For instance, although the potential for big data machine learning is widely recognized, many plants are unable to autonomously access and retrieve data from their Historian databases without the consent and support of their vendors. Digitalization will never happen without access to clean, real-time factory machine data.

Similarly, the Digital Twin concept of GE Predix and Siemens MindSphere is clearly a compelling and valuable addition to the Smart Factory. For a plant owner to access a completely accurate virtual replica of a physical machine in real time is for many the Holy Grail of Industrial IoT.

However, the reality is much more complex.  The Digital Twin remains an elusive big ticket item that is beyond the reach of the most industrial facilities for two important reasons.  First, plant owners cannot justify the price from a TCO perspective.  The Digital Twin is a platform that requires payments to a growing ecosystem of solution providers and it comes at a high cost.  Perhaps more importantly, the deployment of the Digital Twin is extremely labor intensive: external consultants, big data scientists, and design technicians must be hired.  Furthermore, a significant commitment of time from facility engineers and technicians is needed on the client side.

Investing in The Smart Factory:  The Incremental Approach

The high cost of emerging technologies and the inability of plants to access the necessary data to operationalize Smart Factory solutions are likely to inhibit the adoption rates that many of the Analysts (and vendors) have forecasted.

The good news is that there are still practical opportunities to add Smart Factory functionalities, albeit incrementally. From our experience engaging at both an executive (plant owner and/or CEO) and operational level, we recommend the following guidelines for investing in Smart Factory technologies:

Tap into your own existing infrastructure and assets

Disruptive change always attracts technology vendors seeking ways to capitalize on new opportunities.  If you are a facility owner, there is a good chance that someone will pitch you on new sensors, new data infrastructure, and new software solutions.  Let’s go back to basics: your factory produces widgets – technology is merely an enabler to make better widgets and at the lowest cost. Since over-investing in new technologies is a risk or unaffordable luxury, plant owners may need to modify or upgrade their existing systems.

In the case of big data machine learning for asset maintenance, it is not economically viable or operationally scalable to rip-and-replace existing systems.  Your Historian database is a sunk cost and already contains massive amounts of data that you are not using.

I believe that the model that Presenso offers factories should be used for other Smart Factory technologies. Presenso provides a cloud-based solution that continuously extracts data from the Historian and analyzes it in the cloud.  There is no need to purchase new hardware or install software at a local level.  The critical element is that Presenso was built to fit into the existing environment and can be installed with no disruption to factory operations.

Plan for limited labor expense growth

One of the overlooked aspects of Smart Factory investments is that they are people intensive.  It is unrealistic to add big data engineers to your facility to support machine learning initiatives.  These people are hard to find and are typically recruited into high paying industries such as consulting and financial services.  Similarly, your facility technicians cannot move from the factory floor maintenance to computer terminal monitoring overnight.  With cloud-based solutions such as Presenso, there is no need to make additional investments in labor overhead because the heavy lifting of the solution was done via remote with the support of expert big data scientists and algorithm engineers.

Cybersecurity is always a wise investment

The threat from state-sponsored and criminal entities continue to grow. The Energetic Bear attack targeting control systems should keep factory owners up at night.  Analysts indicate that the attackers used basic attack methods to infiltrate SCADA systems such as Metasploit and the Havex Trojan.  Both of these are easily accessible and commonly used by hackers.

The binary choice of cloud versus on-premises is no longer relevant. No system is impenetrable: the notion that the cloud is vulnerable to third party attack is no more valid than the argument that on-premise is inherently more secure.  With cybersecurity, an entire organization is only as strong as its weakest link.  For example, even if one accepts the position that GE’s seemingly endless investment in cybersecurity should alleviate factories’ security concerns, a junior level CAD technician can misplace or “monetize” plant blueprints.

As the threat of cyber-attack grows and new forms of industrial espionage emerge, solutions that can detect threats and data breaches should be prioritized.

Small, Incremental and Nimble

The Smart Factory requires a new mindset that is flexible and opportunistic.  Over time, new investments in technology are necessary.  Some of the skills of current facility staff may become redundant and new job functions will emerge.

Plant owners need to balance short-term operational goals with long-term strategies. Investment in new solutions is a risk, especially when larger vendors are not yet able to provide transparency to their product roadmap.

In conclusion, my advice is to explore and pilot technologies that tap into existing data sources and machine learning that does not require additional hardware or software.  Not all change can (or should) be disruptive.

 

Like this article? Please share it using these icons....Share on Facebook
Facebook
Tweet about this on Twitter
Twitter
Share on LinkedIn
Linkedin
Avi Nowitz

Avi Nowitz

Avi Nowitz writes about the financial impact of Machine Learning on the industrial sector. Avi started his career on Wall Street as an equity analyst, providing institutional investor research on manufacturing companies. After completing his MBA in Finance from New York University, he joined the consulting division of Peppers and Rogers Group where his international clients included Bertelsmann, Ford Motor Company and Doğuş Holdings. For more than a decade, Avi led the Microsoft Consulting Practice at New Age Media and managed the execution of initiatives in EMEA and APAC.