Acting on data is now key to affecting last-mile optimization of production. However, reskilling plant personnel for digital maturity and securing their buy-in for data-driven solutions is a journey.

Our experience with manufacturers has shown us that throughout this journey, establishing customer familiarity with (and trust in) actioning the insights of data science and AI modeling is paramount. And yet, herein lies a caveat: it is also simultaneously essential to preserve and utilize the ‘pre-digital’ plant wisdom around the targeted manufacturing process.


An AI-as-a-Service (AIaaS) model ensures that each phase in the digital adoption journey is contextualized and thus adaptive to dynamic feedback. An AIaaS approach allows for the technology and the processes associated with implementing it to flex according to the emerging needs and ambitions of the plant. In the final analysis, collaborative value tracking and measurement should be implemented conscientiously and in collaboration between the manufacturer and AI partner. Setting targets for value uplift and the means to measure it eliminates the implied risks of adopting a data-driven solution for the plant.


Hyram Seretta, DataProphet’s Head of Growth and New Markets is often the first port of call for manufacturers who are considering our AI use cases. A manufacturing industry professional with a Masters in Electronic Engineering, he specialized in ship target recognition using neural networks. Not surprisingly, Hyram is passionate about the potential of AI to redefine continuous improvement of production. The predominant part of his career prior to DataProphet had been in supply chain, specifically in planning, and then specifically in manufacturing where he ran five malting plants across Africa. Prior to this, he had run a prime and standby power solution provider with a manufacturing facility—producing generators as large as 2500 kVA. As Hyram recalls: 

“The change management that we went through was to bring about an approach broadly called Manufacturing Excellence. Manufacturing Excellence is exactly what it says. It’s about building capability and a way of doing things to secure a culture of continuous performance improvement whilst ensuring its sustainability. The byproducts, amongst others, are reduced costs, improved efficiencies, and improved quality and service. Now, only some of that was digital. But the change management involved, in our case, kicked in because when you build a new plant you have to perform operational readiness at that new plant. In some cases this means existing staff need to start executing new practices; in others, it might be new staff learning established systems. Either way, even as the plant is being built and commissioned, you have the requirement to produce as much yield as you can at the highest level of quality at the lowest cost. So you’re commissioning while you’re running.”

Hyram goes on to explain that in order for a new plant to achieve a state of operational readiness, it may be a matter of urgency for managers to change people’s behavior from what had been customary. This would be the case with new personnel for whom Manufacturing Excellence work practices were not the norm and assuming this is the capability and culture one wants to build. Irrespective of the particular system, Hyram points out that Manufacturing Excellence is a journey with many variables but one goal—a facility that runs with the right cadence and the right rhythm to build sustainability. He also emphasizes that achieving that in a plant can be a massive change management challenge where manufacturing best practices are not being implemented.

In today’s data-centric context, advanced industrials must factor in an extra dimension of change capability; Hyram sees merging operational technology (OT) with IT as the next defining element of excellence in manufacturing:

“In order to bolster a culture of continuous improvement in our digital era, you need manufacturing and operational excellence principles that utilize leading-edge digital-operational technology. So, in the same way that all the other best practice pillars need to be implemented for continuous improvement, so does OT need to augment operational excellence practices.”

Successful industrial change management processes start with a sense of urgency centered around a common ideal and a common strategy. Hyram recounts that in his plant, with 100 to 150 personnel, a vision about what sustained continuous improvement could deliver had to be communicated properly and consistently across the plant. Focused, committed vision empowers and incentivizes stakeholders. Hyram categorizes a change journey that, over time, makes the facility-defined project of Manufacturing Excellence a day-to-day reality. Similarly, in the new era, context and urgency need to be provided to plant teams—as far as incorporating data-driven technology as it relates to OT and IT systems.


The critical use case Hyram draws on to establish urgency for a data-driven solution is around root cause analysis (RCA). In manufacturing, RCA is the traditional (and necessary) approach to a missed key performance indicator (KPI). It is followed by a corrective action request (CAR) and an engineering change note (ECN). However, due to disparate data being drawn on from personnel in different teams and the limitations of human analysis, this closed-loop process of finding the root cause and fixing it can be extremely slow, laborious, and exasperating. The result is fire-fighting instead of continuous improvement, which can weigh heavily on morale.

Hyram categorizes this challenge—all too familiar to cross-functional plant teams performing an RCA—as a great window of opportunity for a mindset shift towards digital maturity. It is an opportunity to appreciate the potential value of an idea like Manufacturing Excellence 4.0:

“If the plant teams have an effective data-driven optimization technology at hand, this can be a real incentive—provided they can keep it in the front of their mind that the technology is an enabler and a value driver. If everyone’s on the same page that the tech is extremely important to produce better yield and quality down the line and will save them hassle, then a smart factory initiative will work.”

DataProphet PRESCRIBE is a case in point for reconceptualizing the RCA phenomenon as a Manufacturing Excellence 4.0 opportunity. PRESCRIBE is a prescriptive optimization solution that:

  1. uses deep learning discovery applied as prescriptive analytics to identify in advance the critical parameters that bear on production KPIs.
  2. embeds successful RCA into an AI-driven update loop. 

With these functionalities, PRESCRIBE can be defined as an Expert Execution System (EES). Preventative action is generated to a human-machine interface (HMI) for operators. Crucially, this happens ahead of the fire-fighting that typically ensues as production is maintained while cross-functional teams react to process anomalies stemming from an unidentified yield detractor.

For plant engineers, process engineers, and operators, PRESCRIBE works as a case of data-driven technology augmenting an RCA process (making it faster, more accurate, more efficient, and easier to do). Parsing process data from the historical operating region where the plant was most stable and optimal along with actual quality data, PRESCRIBE eliminates many of the variables plant teams agonize over in trying to determine missed KPIs. Beyond this, it also establishes a system where the traditionally executed RCA is avoided in the future.


In sum, it is crucial for plant teams and manufacturing executives to bear the following in mind—the data-driven continuous improvement journey provided by AIaaS integrates with existing operational best work practices. In other words, established continuous improvement standards remain essential pillars.

For this reason, conceptual rebooting for smart factory initiatives does not mean seeing an AI-driven plant as a threat to its current Manufacturing Excellence regime. On the contrary, Manufacturing Excellence 4.0 is reframed as an augmenter and a force multiplier. It uses leading-edge OT and data science techniques to drive and measure additional production value. Along the digital maturity curve, technology captures this new value while seamlessly integrating with a factory’s fundamentally successful, pre-digital, standard operating procedures.

%d bloggers like this: