A “data-driven” operation is defined as one that harnesses its operational data to augment the experience and judgement of operatives, managers, analysts and engineers as they plan, organize, conduct and control their processes.
Note that almost all strategies for data-drivenness are not high-tech or new-tech as much as they are modern-day knowledge, skills and software in action. The words “almost all” are used to allow for the small specialized domain of IIoT-based strategies which entail high-tech and deep capital. The former rather than latter are the subject of this website; leaving IIoT to the purveyors of its technologies and engineering.
Every firm’s current-day challenge is to find the actionable training to modernize the role holders in their operational processes. The greatest challenge is to find training without conflict of interest; meaning that the content is not self-serving to the its source. Instead, with the content, the operation can proceed, self-directed, without the offered services or products of the presenting instructor or enterprise.
The training sessions of this page meet that standard. They were designed by an operational SME who has learned data science and thinks in terms of how to explain the possibilities to other SMEs. This can be confirmed by inspecting the session slides while also confirming that the session would be new knowledge to the organization.
Six interrelated training sessions are structured with one session as a framework and five others as deep-dives into sections of the framework. The sessions and links to each are as follows:
The Framework for Data-Driven Operations: The first step to becoming a fully data-driven operation is that process role holders must reach a clear, implementable understanding of data-drivenness. The purpose of the training session is to take the first step. The session is framed upon the critical-mass of threshold set of knowledge, skills and software that must be in place to be fully, effectively and efficiently data-driven with which data-drivenness can be built up from the grassroots starting with units as small as an individual taking an initiative within their personal autonomy.
Build Super Tables from Operational Data: From normal functioning, massive troves of data are captured in thousands of tables in the background of a firm’s operating systems such as a CMMS. To be capable of building a data-driven operation, role holders across the operation must know how to extract tables from their systems and join and mold them into single “super” tables. The purpose of the session is to teach how to build super tables demonstrated by techniques that will typically apply to everything envisioned by a builder for their own purposes.
“R” in Action: Hands-On: “R” is a non-proprietary open-system software that is as powerful as the best of commercial offerings. With it, all analytics to data-drivenness are conducted. To give a perspective for how to download and work with “R” the slides as an example will show how to work through a case such that the user can have their first hands-on experience with the software. From the experience it will become apparent that its users work with go-by examples rather than code analytics from scratch—know the analytics, not the code.
Modeled Insight Deliverables in Data-Driven Operations: One of four types insight deliverables to data-driven operations are “modeled insight.” They apply the methods of machine learning and artificial intelligence to ask and answer five types of questions that are not otherwise possible —relationship, difference, time series, duration and apparency. In contrast to the framework session, this session expands upon the many variations to the questions that are inherent to the various models.
Rechart Operational Processes to be Data-Driven: The training session, as a case, works top-down through the steps to rethink the management processes of a maintenance operation through the lens of what data and analytics make possible. The sequence applies to any operation but uses maintenance as the case. At the top is to establish how the firm, through the plant, competes and wins in its industry and how the competition is scored financially. Next, is to establish a proxy to the top measure of competitiveness by which the maintenance operation would be measured in the same financial terms as the grand score. Thence, is to establish the set of operational scenarios that would constitute maintenance performance at the pinnacle of the proxy. These constitute the North Star to the maintenance operation. Finally, is to establish and chart the structure of maintenance management processes that are required for the plant to be able to operate at the pinnacle.
Project Plan for Implementation at the Grassroots: An organization will need an implementation plan to put the knowledge, skills and methodologies of data-drivenness in play. This requires a path of stages, steps, tasks and deliverables along which to steer each chosen operation from its current state to being fully data-driven. This session is generalized—not intended to fit any particular case—so that you can see what needs to happen and modify the plan to fit the nature of your organization. Most importantly, the plan is structured to cause the transfer of knowledge, skills and methodologies to the role holders in implementation and subsequent functioning.