Troubleshooting Unusual Fault Conditions

Best approach sometimes takes more than simply identifying and resolving the problem.

Years ago, when I was actively involved in troubleshooting — that is, in the remedy of unusual fault conditions — a number of these led me to reflect on best how to approach this, not only in relation to the best ways to identify and resolve faults, but in relation to our approach to reports of lift malfunction, and how we might learn to better understand and evaluate witness evidence. My experience of this has proven particularly useful in relation to forensic engineering and expert witness work, in which I have observed first-hand the way in which the approach of an investigator may be unduly influenced, and how, subsequently, the progress of the investigation is distorted and disrupted by erroneous and/or misinterpreted witness reports and the detrimental effect this exerts upon an investigation.

Repetition of Lift’s Odd Behaviour

An experience early in my career involved a hydraulic lift installed in an aged-persons home. The lift was an eight-person, suspended hydraulic installation, which incorporated a jigger-type hydraulic jack arrangement and served two floors.

For several years, reports had been received of the lift ascending with the car and landing doors open. These reports were repeated periodically and, despite quite thorough investigations, no repetition of the reported event was observed, nor was any fault found, even after extensive testing, simulation, and extended periods of operational observation. As a result, a tendency to dismiss further reports developed, with these being attributed to the ramblings of allegedly bored or senile residents of the home.

On the occasion in question, the reports had been verified by several residents, such that it was decided to investigate once again. I attended with a colleague who was well versed in resolving control problems. We carried out numerous tests and simulations, checking circuits and wiring connections, but were unable to reproduce any indication of the fault reported. After five or six uneventful hours, we finally gave up and packed up our tools.

As we were leaving, an elderly gentleman in the building lobby said, “Ah! You’re the fellows sorting out the lift problem?” My colleague, who by this time was suffering the frustration experienced by lift engineers attempting to detect and resolve elusive control problems, was somewhat sharp in his response. The man was a little taken aback and responded, saying, “Do you want me to show you?”

Now, the effect of such a statement is to make one stop and listen. We said, “Yes, if you can.” He said, “Watch,” and called the lift and entered the lift car which, as he pressed the call button, began to ascend with its doors open, right there in front of us. We were astounded.

Having ensured the immediate situation was safe, we asked how he had made this happen. He explained and demonstrated that, by simultaneously depressing the first-floor car push button and the door-open push button, the lift car would start to ascend with the doors open.

Suitably admonished and apologetic, we expressed our thanks and headed back to the motor room. How could this happen?

Checks did not reveal any erroneous or undue electrical connection or contact between conductors and/or components of the door-open button circuits and the up-direction circuits, which may have operated to energise the up-direction contactors.

In the absence of any visible deficiency, we looked at the electrical schematics in closer detail. The controller was a 1970s relay logic design. After spending some time analysing the circuits and controller, we were unable to find anything obviously wrong or misconnected. However, the lift consistently ascended with open doors whenever the first floor and door-open push buttons were depressed simultaneously.

Further analysis led us to a blocking diode which formed part of the door circuit. The circuitry was quite complex, and interconnected in a number of areas, and with elements of switching on the negative side.

We decided not to look at the hardware, but to focus purely on the schematics, eventually noting that the blocking diode appeared to operate in order to prevent a possible sneak-feed (a term prevalent in those days in relation to complex relay-based control systems) from the door-open circuits to the up-direction circuits.

Using a meter, we tested the function of the diode (the visual appearance of which was fine), and found this to be short-circuit, and in conducting mode. This was unusual in that such diodes were understood to fail to an open-circuit condition. Further inspection revealed the problem: Someone had inserted a tiny section of thin fuse wire (5-6 mm in length) behind the diode (which was a miniature device mounted in a block of around 20 similar diodes) and in the rear of the control panel. At some time in the past, the diode had failed (open circuit) and a service engineer who hadn’t recognised the significance of the diode had shorted it out, no doubt restoring the circuit to operation, but unwittingly introducing the fault.

Never Dismiss Witness Evidence

So, what do I take from the experience? Firstly, never, without proper consideration, dismiss witness evidence, no matter how seemingly incredulous this may be. Second, what of the original circuit design? Should the failure of a diode costing a few pence be such as to give rise to a dangerous situation? Should circuits be designed in this way? Could the diode and circuit have been a late design addition to correct an earlier oversight by the controls designer? What of the action taken by the service engineer who shorted-out the diode? I must assume that he (it would invariably have been a “he” back in those days) was unaware of the possible consequences of his action. However, if it was decided that a short was to be applied (albeit that it clearly shouldn’t have been), should this not have been applied in a way that was obvious such as to ensure early removal and replacement of the diode?

A number of similar experiences over the years have led me to develop an inherent resistance to loose-ends in failure investigations. I recall one particular case, in which the evidence of two key witnesses of fact was disregarded in favour of a technical explanation proffered by a lift contractor manager, who was not even present at the time, but which suited the situation and provided a simple and convenient explanation. The erroneous and misleading explanation was accepted by an enforcement authority and used to support the issue of a prohibition notice. Subsequently, I was able to prove that the technical explanation was in fact impossible, and that the evidence of the witnesses, no matter how problematic or inconvenient, was factual and accurately reflected what they had observed in terms of the events which had arisen at the time, and which they had witnessed first-hand.

I also reflect upon the fault-finding skills required to work on the complex relay logic systems and floor selectors which prevailed prior to the microprocessor. I had the pleasure to work with a number of engineers whose troubleshooting skills were outstanding. In those days, it was widely asserted that an individual either possessed or did not possess the skills necessary. In effect, these skills were widely considered to be inherent to the individual and to be something that could not be readily taught unless the individual possessed the inherent traits. I guess in terms of human psychology and trait theory, there might be something in this assertion, although the reality almost certainly lies more in the approach adopted by the individual, which should be logical and based upon a sound knowledge of the equipment and founded in the underlying engineering principles. One excellent theory promoted within Otis related the process to that of asking the right questions, with the quality of the question asked being determinate of the speed of resolution, with poor quality or irrelevant questions extending the process by leading the troubleshooter “up the garden path,” so to speak. And, even worse, with poor answers further extending the process and frustrating the troubleshooter.

Witness evidence (suitably tempered in terms of interpretation based upon informed theory and knowledge, but never wholly ignored) was considered important in terms of establishing a “story” of what had occurred. A great emphasis was placed upon the application of the human senses in observing and listening and in the sense of smell (who will not recognise the smell associated with a dragging brake lining or a deficient plate-rectifier, each of which are familiar to the experienced troubleshooter?) and in a developing a “feel” for the equipment (ride quality, switch function, rope tension, linkage operation, etc). One of the skills I was asked to develop during my legal training was that of observation – how to properly look at something and to actually see what’s there, as opposed to being blinded by preconceived perceptions. This may sound ridiculously simple or even strange, but I have tested this proposition over the years on a number of colleagues and I never cease to be amazed at how readily we are able to see what we expect to see, rather than what is actually present.

Many problems which affected the larger relay-based systems were accumulations of previous wrong fixes, the effects of which could accumulate to become so complex that provision was made for readjustment of the complete system every few years, when the overall effect of these erroneous adjustments and fixes was such as to render a fault-finding, process-based approach impractical. It was necessary to start again at the beginning.

I once spent several weeks with a colleague readjusting a 1950s group control system with gearless Ward-Leonard drives. This, with its numerous controllers and a system of dispatchers, together with mechanical floor selectors, presented a real challenge, requiring that we should develop an understanding of the control approach which prevailed in the 1950s (which sounds simpler than it is), and then remove the numerous erroneous “fixes” that had been applied over the previous 25 years (when engineers had addressed symptoms, rather than underlying causes), and finally get the system up and operating. In addition, we were tasked with installing circuitry which would allow two of the cars to be operated separately of the group, as a duplex, at certain times during the day (in effect, two separate groups serving a common lobby but different destination floors), and to revert to full group system control each morning and evening. Few of today’s lift engineers will be afforded the opportunity to work with such systems, and I still appreciate the skills I was able to develop from the experience — albeit that this was at times frustrating.

The modern troubleshooter has the advantage of the sophisticated, and comparatively cheap, test equipment available nowadays, although an ability to use and apply this equipment effectively remains critical.

Most Important Tool: The Human Mind

And, finally, the key skill relates to the use of the most important tool, the human mind. In our minds, we constantly evaluate and reevaluate evidence, processing this and relating back to the questions and/or formulating new questions, always assessing and reassessing the evidence. However, the frustrations associated with troubleshooting, which are exacerbated by time and cost pressures and by other pressing assignments, must be recognised and managed. Indeed, I well recall the times I have seen engineers, through sheer frustration, throw tools at a motor room wall (I guess this may be considered a form of management?). I also wonder whether a level of frustration may underlie, or contribute to, the application of the symptoms-based “fixes” as engineers become too frustrated to seek out root causes and are led to do whatever will fix the immediate or apparent problem.

The Otis engineer behind this analysis of troubleshooting considered that an hour onsite should usually be sufficient to produce enough “yes” or “no” answers to identify the likely issue and that, if this is not the case, then the likelihood is that the right questions are not being asked, and the approach should be reevaluated. The same engineer considered 95% of calls to be very simple, a further 3% to be misleading and the remaining 2% to be very difficult, and recognised that troubleshooting is fun as it presents a challenge and, when successful, rewards the ego.

 In terms of evidence, we are prone to accept explanations of events which suit a particular perspective, a set of circumstances and the context of the time. Once these stories become established and internalised, we become reluctant to accept alternative explanations, regardless of how rational and convincing these may be. We become unduly influenced by the extent to which our own erroneous — or even irrational and unsupportable — explanations have become internalised, often so deeply that our thought becomes entrenched and immovable. The internal psychological conflict is so uncomfortable that we become prone to adopting a defensive “don’t confuse me with the facts” approach.

A number of the skills inherent to the effective elevator troubleshooter overlap significantly with those of the effective accident/failure investigator, forensic engineer and/or expert witness, and I still consider the time I spent working in this area, and with the complex relay logic equipment of the time, to be invaluable to this day.