Jane M Fraser

What’s new?

Modern Machine Shop Online reported on 18 September that Rigibore’s ActiveEdge wireless boring bars (working in conjunction with sensors) automatically adjust the boring system to reliably produce holes within the desired tolerances for the hole diameter. The automatic nature of this system eliminates any errors due to human adjustments.

What does it mean?

Many parts require a hole to be drilled; the hole must be within a specified tolerance range, that is, the diameter must be greater than a specified lower limit and must be less than a specified upper limit. Reliably drilling holes within the tolerance is necessary for the part to function properly. With an in-process measurement system, each hole is immediately and automatically measured. The measurement is recorded and, if it is out of specified warning levels (that are within the tolerance range) the drilling system is adjusted. The warning levels mean that the most recently produced part is still within tolerance (and the part does not need to be rejected), but the part is nearing the upper or lower tolerance limit.  With the Rigibore system, that adjustment is done automatically; the measurement is sent wirelessly to the cutting system, which automatically makes the correction.

This system of measurement and adjustment is a negative feedback system. If the diameter is too large, the cutting diameter is made smaller for the next part; if the diameter is too small, the cutting diameter is made larger. A thermostat is the same type of system: if the room is too warm, the temperature is reduced by cooling; if the room is too cold, the temperature is increased by heating. You do the same when you drive: if your car is drifting right, you make an adjustment in the steering to head to the left; if your car is drifting left, you make an adjust to head to the right. This feedback is called negative feedback because the system is adjusted in the opposite direction of the detected measurement. A negative feedback system, if well designed, keeps a goal within specified limits.

The article reminds us that accurate measurement is necessary. The article quotes Rigibore’s president as saying “Gage accuracy has to be rock solid.” If your measurement says the room is too hot, but it actually isn’t, the correction made to cool the room may actually cause it to become too cold.

The article reminds us that randomness can occur; as mentioned in the article, a measurement might be inaccurate due to a piece of dirt in the hole. The Rigibore system makes an adjustment only after detecting two measurements outside the warning limits. Feedback systems, if poorly designed, can chase random effects, creating more rather than less variation. Dr Deming’s funnel experiment was designed to illustrate the ill effects of tampering with a system.

The article reminds us that the timing and the amount of adjustment have to be selected carefully. Rigibore has designed the system to make adjustments without increasing the time it takes to bore a hole. The system is also designed to make very small corrections. Overcorrection in driving is well known to lead to accidents.

While the article doesn’t remind us, feedback systems work only if the system being controlled is well understood. The goal must be clearly stated (a hole with diameter within tolerances), the measurement to be made must be clearly defined (the diameter), the measurement system must be accurate (“Gage accuracy has to be rock solid.”), and the action to be taken to correct the system must be well understood (if the diameter is too large, reduce the size of the bore by a small amount).

Negative feedback control is a simple model but its concepts underlie much of engineering. Control charts, one of my areas of expertise, are based on the same concept of monitoring a key performance variable and taking corrective action when measurement indicates the system has deviated from ideal settings. Control charts add the interesting concept of also monitoring and controlling the variation in that key variable.  Reducing variation is a valuable strategy for keeping the variable within limits.

What does it mean for you?

The most important part of negative feedback control is that it is closed control, that is, the system itself is measured before a decision is made on what control to implement. That simple concept may need reinforcing with some who think they can steer without first finding out where they are currently heading.

If an the organization identifies a goal to be kept within certain limits (for example, growth rate in sales, the number of product design changes being implemented, or the amount of emails and other communication being sent in the organization), the organization’s leader can ask for a clear statement of the goal, a clear definition of the measurement, an accurate measurement system, and – the most difficult part – a clear description of the system for controlling the goal. What are the factors that can be used to heat up the room (increase sales) or to cool the room (decrease sales)? What is the time scale on which each of these factors operate? What are the levers that the organization can use to affect those factors? No organization can probably answer all those questions, but the discipline of asking them focuses discussion and research.

The need to have that clear understanding supports the idea that the role of leadership is to study and understand the system that is their organization and the system within which the organization works. Workers work in the system; management works on the system.

The metaphor of control in this simple model may lead some to think that the lesson is top-down management, but the Rigibore system is so successful because the measurement and corrective action have been delegated to the lowest level of control, the actual tool. The people and sensors closest to a key variable must have the authority to measure and control that variable.

Finally, this MMS Online article on Rigibore highlights the importance of definition and measurement. You cannot control what you can’t define and measure. See this explanation from the US Bureau of Labor Statistics on the definition of the unemployed and the employed and, while they are correct to state that the basic concepts are simple, the details are devilish. The BLS itself reports the values of several measurements not just of unemployment but also of underemployment.

Where can you learn more?

The mathematics of negative feedback control is simple and elegant, but also leads to more complicated models. See, for example, this explanation from chemical engineering. Then there is the textbook used for the course in control systems at my former university (Colorado State University-Pueblo), Feedback Control Systems by Charles L Phillips and John Parr, now in its 5^th edition.

Feedback control goes by other names. Deming describes balancing loops. Senge’s fifth discipline is systems thinking and his first systems archetype is balancing feedback.

Metaphors are powerful. Contemplating their use can lead one down numerous deep rabbit holes. The Stanford Encyclopedia of Philosophy says: metaphor “has attracted more philosophical interest and provoked more philosophical controversy than any of the other traditionally recognized figures of speech.” Gareth Morgan is the best known proponent of the use of metaphors to understand management, as described in his book Images of Organization, first published in 1986 and revised in 2006.

This work is licensed under a Creative Commons Attribution 4.0 International License.