What are the hurdles to Real-Time Operational Excellence?

I see a significant increase in “operational Transformable projects” , but too often it is talk, or dreaming, and when we discuss the ideas people like, but they really miss the challenge. Too often they fall back into the traditional approaches can we get access to information, through reports and dashboards. 

Born out of the frustration to gain the transparency to “what is going on NOW”. Yes it is a journey for “operational excellence “ and it will not be done once or ever over in this ever “speeding , agile world”.

Taking a step back and understanding the hurdles to getting to “Managing by Exception”. I thought the image below simplified the discussion.

Understand where you are, and set a vision of where you want to be, and this goes back to shift towards “activities” design vs application or even role.

Above you can see how not having the data in context, or even accessible is key, this is seen in the two bottom challengers. As one customer said last week, how do eliminate cleaning data every 3 months. The answer is simple, capture data as close to the source, validate and structure it as close to source as possible, so now you are storing valuable, trusted information, and you can depend upon it.
But now you have the information people put it into reports, and dashboards, for decisions to be made, but did it get to correct person, did it get decided upon in timely manner, why it did not escalated, or collaborated to accelerate the decision. The system must provide this framework for escalation, and ability ask/ share.

With the changing roles, and people on plants, and the horizontal structure, do we know the decision was made, “accountability” is important when something is sent. Too often tradition alarms, notifications have no accountability, the only way a team works is that they understand their role, and responsibility for decisions.

Then you come to final hurdle “what do I do having made the decision”? This needs to consistent processes across different workers of different experience. Also the system has to shift to a “crowd sourcing” culture of continuous improvement and everyone is empowered to contribute.

This may seem so simple but it is fundamental to the “transformation in Work” yet so many programs are missing these basics.

I will follow this up next week again on why “People and Processes” are key to take the automation to the next level.

We need to improve the speed and accuracy of big data analysis in order for IoT to live up to its promise!

I was listening and reading the debate on IOT, and this article was layered with good amount of reality.

“As the Internet of Things (IoT) continues its run as one of the most popular technology buzzwords of the year, the discussion has turned from what it is, to how to drive value from it, to the tactical: how to make it work.

We need to improve the speed and accuracy of big data analysis in order for IoT to live up to its promise. If we don’t, the consequences could be disastrous and could range from the annoying – like home appliances that don’t work together as advertised – to the life-threatening – pacemakers malfunctioning or hundred car pileups.”

http://www.wired.com/2014/11/iot-wont-work-without-artificial-intelligence/

This follows on from my discussion 2 weeks ago around the need to avoid just gathering data, vs gaining the proportional amount of knowledge and wisdom, which brings in a term you hear a lot “machine learning”.

“Wikipedia defines machine learning as “a subfield of computer science (CS) and artificial intelligence (AI) that deals with the construction and study of systems that can learn from data, rather than follow only explicitly programmed instructions.”

“The realization of IoT depends on being able to gain the insights hidden in the vast and growing seas of data available. Since current approaches don’t scale to IoT volumes, the future realization of IoT’s promise is dependent on machine learning to find the patterns, correlations and anomalies that have the potential of enabling improvements in almost every facet of our daily lives.”

In the industrial world this more applicable than nearly all industries, and in many cases we are already applying “machine levels” at different levels. A key part in the shift from “Information” to “knowledge” is having the tools to drill into historians based on events and discover learnings and patterns. Once validated and discovered these are turned into “self-monitoring” conditions to understand the current state of the device, and predict / recognize conditions well before they happen. Providing the “insight” to make awareness and decisions where the machines/ devices are telling you where the opportunities are. But a key part of machine learning is that this knowledge in not a once off step, it is a continuous evolution leveraging the gathering history data and developing increased amounts of knowledge.

The next step is to then apply proven or recommended operational processes to these decisions, so as a condition is recognized by the devices, either they take an action automatically or they recommend the action to the user in a timely manner with escalation. A key transformation IoT brings is the increased speed at which trustworthy knowledge is made available for actionable decisions to taken.
I like this phrase:


 “It’s time to let the machines point out where the opportunities truly are.”

Trustworthy Operations Management Solutions

I asked Stan to contribute a blog on a topic that he and I are asked, that of "trust worth systems/ data" this is an incredible critical item as we move to "actionable decisions"

Blog by Stan DeVries.

When younger workers are asked about how “trustworthy” solutions should perform, a common response is “it just works”.  This is a reasonable but demanding expectation, and it is a combination of availability, accuracy and acceptable user experience in all facets.  One aspect of operations management solutions which makes this expectation more challenging is that these solutions are inherently more complex – they include at least 2 software applications, sometimes 15 or more.  And complexity tends to reduce availability.

Several customers have asked how to practically achieve and sustain “trustworthy” operations management solutions.  An appropriate analogy is a fuel gauge in a car; if it is functioning less than 100% of the time, users won’t trust it at all.  The following are best practices:

• Design the solution to automatically handle many failure modes, including user error.  Most of the design of automatic teller machines (ATM’s) is handling failure modes.  Methods include automated workflow for missing or grossly erroneous data, software and machine health, network outages etc.

• Design the solution for some redundancy, including “store and forward” of data to withstand network outages and other failures.  Note that this technique is only usable when the software applications can rapidly process the restored data while processing “new” data.

•  Design the calculations for sufficient accuracy and availability.  Simple mathematics is much more available, but much less accurate, than complex mathematics.  Technology is available that delivers high accuracy and has built-in logic and knowledge to overcome many failure modes including “solver” errors, sensitivity to missing or inaccurate input data etc.

• Design the solution’s outputs using the “4 rights” instead of the “4 anys”:

1.  Information should be delivered at the “right” time (which might be earlier than “real time”) depending upon the operations management conditions.
2.   Information should be delivered to the “right” persons.  Operations management solutions tend to broadcast information including undesired performance and tend to broadcast information which is irrelevant to most users, which means that users must filter out information that seems like “spam” and users must learn to trust the solution.
3.    Information should be delivered in the “right” context.  There is an analogy which characterizes “data”, “information”, “knowledge” and “wisdom”, where “data” is raw data, “information” is trustworthy data (may include substitutions and reconciliation), “knowledge” presents a comparison of information to targets, constraints and similar information, and “wisdom” is prescriptive instructions to exploit desired opportunities and to prevent or minimize undesired conditions.

An operation management solution evolves technology is introduced, the operation evolves and as users increase their dependency and trust in the solution; the above methods are good fundamentals for the solution’s lifecycle.