Expec­ta­ti­ons and tech­ni­cal limits

This is necessa­ry to be able to use mea­su­ring instru­ments sen­si­b­ly.

A new high-qua­li­ty mea­su­ring device …

… used to attract a lot of curio­si­ty and respect. One hard­ly dar­ed to turn one of the con­trol knobs – it could have been easi­ly dama­ged.

Today we have – through con­stant pre­sence – a much more rela­xed rela­ti­ons­hip to tech­ni­cal equip­ment. There’s an app for ever­ything : down­load, run, done. If this does not work right away, chan­ges can be easi­ly revo­ked. Usual­ly, the­re are no serious con­se­quen­ces to fear.

As a result of this pro­cess, the expec­ta­ti­ons of the tech­no­lo­gy have also chan­ged. Devices must be robust, easy to use and under­stand wit­hout much know­ledge.
If I don’t under­stand some­thing right now, then I need ano­t­her app that sol­ves the pro­blem for me. Howe­ver, we do not want to deal with the back­ground to a pro­blem. Chan­ging some­thing manu­al­ly or spen­ding more than a few minu­tes dealing with a topic is no lon­ger on the agen­da.

That’s how you get care­free first,
later unsu­spec­ting and final­ly reck­less.

This is dif­fe­rent in indus­try.

Here, basic know­ledge and the desi­re for some­thing new is still in demand and necessa­ry – both for set­ting up new sys­tems and for main­tai­ning or repai­ring exis­ting sys­tems and plants.
Sys­te­ma­tic net­wor­king has reached all are­as of tech­no­lo­gy. But this also means that more and more peop­le have to deal with the issue.

The con­se­quen­ces have a mas­si­ve influ­ence on ent­i­re occupa­tio­nal fields. A machi­ne fit­ter can’t work any lon­ger wit­hout a lap­top. Mean­while, he also needs spe­cial mea­su­ring instru­ments to check his machi­nes and to find errors suc­cess­ful­ly. No 32 mm open-end wrench helps and with the mul­ti­me­ter alo­ne you won’t get very far.

Unfor­tu­n­a­te­ly, the expe­ri­ence of net­wor­king from the pri­va­te envi­ron­ment and the resul­ting expec­ta­ti­ons regar­ding net­wor­king in auto­ma­ti­on now stand in the way of our machi­ne fit­ter. He thinks a spe­cial device can sol­ve his pro­blems in minu­tes. He expec­ts a clear indi­ca­ti­on of the exis­ting pro­blem wit­hout having to think about it fur­ther.

Unfor­tu­n­a­te­ly, we have to disap­point our mecha­nic at this point.

Screen­shot You­tube

Let’s take a litt­le men­tal trip to the gara­ge. The­re the mecha­nics now know from pain­ful expe­ri­ence that an error mes­sa­ge on the dis­play of their OBD dia­gnostic device is just a snapshot of the past. “A “lamb­da pro­be fault” does not mean that repla­cing the pro­be will sol­ve the pro­blem – even if this is done in 90% of cases for profit’s sake. Many other cau­ses can be con­si­de­red, for examp­le, a loo­se hose over which secon­da­ry air is sucked in.

To under­stand this, one has to deal with the func­tion of the pro­be and the con­trol of the motor. The OBD dia­gnostic device is more or less only able to read the error memo­ry of the con­trol­ler, not to inves­ti­ga­te the cau­se.

Wrong expec­ta­ti­ons

Unfor­tu­n­a­te­ly, it hap­pens that the users of our dia­gnostic devices also approach the mat­ter with the wrong expec­ta­ti­ons. The new expen­si­ve device is hard­ly noti­ced and put asi­de. Once the day of the first use is here, a work inst­ruc­tion is expec­ted direc­t­ly on the dis­play as to what to do to reme­dy the spo­ra­dic standstill of the machi­ne.

Don’t misun­derstand us, we would love to build such a device, but it will not come to that. The sys­tems are far too dif­fe­rent to be able to do such things. And also becau­se the mea­su­ring devices can read much more than just error memo­ry, exten­si­ve trai­ning is necessa­ry if the device is to be used sen­si­b­ly. Our tech­ni­ci­an or mecha­nic will not be able to do much with the mea­su­ring instru­ment unless he is wil­ling to invest time for lear­ning the basics, time for lear­ning the pos­si­bi­li­ties and limi­ta­ti­ons of the mea­su­ring instru­ments and time for prac­ti­cal mea­su­rement and tes­ting.

Our tech­ni­ci­an won’t be able to do much with the device.
if he’s not wil­ling to invest time.

can training flipchart

Basics : how does the CAN bus work, how are data exch­an­ged, what inter­fe­res with com­mu­ni­ca­ti­on, what does a good struc­tu­re look like and which errors should be avoi­ded.

Second­ly, you have to deal with the mea­su­ring device its­elf. What can the meter actual­ly mea­su­re ? How do I ope­ra­te the device ? Which work steps do make sen­se ? Which rea­dings do I expect in a good bus or in a bad one ? The issue of “What can’t I mea­su­re with it?” is also important.

Then it’s time to try out what you’ve lear­ned : First on a working bus : does the meter show the expec­ted data ? What hap­pens when I make small chan­ges ? How does the meter inter­pret this ?

AI – arti­fi­ci­al intel­li­gence – is cur­r­ent­ly on everyone’s lips. But what does AI only do ? It stores what is nor­mal by fee­ding the sys­tem mas­ses of data. Wit­hout a com­pa­ra­ti­ve value, it is hard to judge whe­ther some­thing is good or bad.

The­re­fo­re, after the first attempts to walk with the new mea­su­ring device is far from over. The task now is to crea­te a data­ba­se with refe­rence values. You run the risk of clas­si­fy­ing a signal as “good” too quick­ly just becau­se you never saw it befo­re and have no data­ba­se as a refe­rence. The­re­fo­re, tech­ni­ci­ans should mea­su­re at as many points in dif­fe­rent ope­ra­ting modes as pos­si­ble on their sys­tems, collect as much data as they can, com­pi­le and store lists of nodes and docu­men­ta­ti­on in a mea­ning­ful way – pre­fer­a­b­ly befo­re a first fail­u­re occurs. For examp­le, it makes litt­le sen­se to mea­su­re at a sin­gle mea­su­ring point in the vehi­cle even though the struc­tu­re has four dif­fe­rent bus seg­ments. Here, at least one point should be mea­su­red at each seg­ment, pre­fer­a­b­ly at the begin­ning and end.

But even with all this pre­pa­ra­ti­on, fin­ding a mista­ke will still not be tri­vi­al. It is now a mat­ter of making mea­ning­ful mea­su­rements and drawing the right con­clu­si­ons in com­pa­ri­son with the old mea­su­rement data. Is the bus dis­tur­bed at all ? Is the bus its­elf a pro­blem or is the pro­blem some­whe­re else ? Ques­ti­ons that are easier to ans­wer by com­pa­ri­son than wit­hout data.

It always helps to base trou­ble­shoo­ting on a sys­te­ma­tic, logi­cal pro­ce­du­re. Unfor­tu­n­a­te­ly, the­re is no magic for­mu­la for this (see com­ple­xi­ty and diver­si­ty of sys­tems). But once you have gone through the above steps, you have the best pre­re­qui­si­tes to sol­ve the pro­blem, find the fault and get your machi­ne up and run­ning again. A tech­ni­ci­an who has taken the­se steps has accu­mu­la­ted a wealth of expe­ri­ence that can­not be taken away from him. And with every use, the trea­su­re grows.

Final ques­ti­on :
What does our tech­ni­ci­an know wit­hout a mea­su­ring device ?

Sub­scri­be to our news­let­ter now.

By cli­cking on the “Sub­scri­be to our news­let­ter” but­ton, you agree that we may use the data you pro­vi­de to send you the news­let­ter. Fur­ther infor­ma­ti­on on data pro­tec­tion can be found in our data pro­tec­tion decla­ra­ti­on.