Automation of Gauge Blocks Calibration in LATU

Interesting presentation. Some questions: Does the vibrations produced by the device (e.g. to hear around minute 8) have any influence on the measurement and/or the block comparator? Do you achieve higher or lower uncertainties than calibrating manually? Or the same? The operator has to surveillance the measurement process or can the system work totally independently without any risk of damaging the gauge blocks?

Hello Alexis. The sound that is heard, similar to a motor, is not from the actuator or any device for automation, but comes from the motor that lifts the probes, existing in the original equipment. Anyway, this motor, although it makes noise, in the video it seems stronger than in reality and does not produce vibrations. Another sound at the end of the movement (like friction) is due to the slow movement of the template over the bench, it does not involve the blocks. Regarding the uncertainty, in theory it should have less uncertainty than the manual process since the metrologist does not bring her arm closer the standards, modifying the temperature. However, the truth is that in comparative tests, no changes were obtained regarding the manual process, neither in results nor in uncertainties. Finally, regarding security, yes, the automation is completely autonomous, it is only required to start in the software and return for the report when finished. For this, some safety routines have been added to the software, for example, if the probes do not reach a safety separation (for example, if there are leaks in the pedal pump), it is not possible to execute movements in the actuator, neither during the workflow nor applying a direct movement command (avoding operator negligence). In such a case, the software would attempt to raise the probes 1 minute, and if this fails, the process is aborted.

Hello, very good work! I wanted to know if they saw the possibility of seeing three digits? I mean, see the application 0.001 µm

Hello, good work, Pablo. First of all, answer these questions in the Workshop please. Which software did you use? Labview? Why not python? Does your software read certificate temp to correct the values measured?

A good job! How do you connect this "software" with emission of calibration certificate?

Hi Pablo, As a colleague who invest a lot of time improving and automating different process, I would like to ask you... a ton of questions actually, feel free to answer those that you consider more relevant during your project, I consider each one a important part during the automation labours. || How much time was necessary to automate the process? || How many people intervened? || What is the effective free-time that users are experiencing after automation? || What are they doing with that free-time? || What was the participation of the final user during the automation process? || What do the user think about the visual part? || How much time do you need to capacitate the new personal for using the software properly? || Have you considered using a visual or an auditive alarm in case nobody were present in the laboratory while the system enters a fail state? || Have you considered a force's sensor, current consumption or times between displacements to let the user know that something is wrong or as a self-monitoring system? || Do you have any security system in case results present irregularities? || What happens with data's system while the system is in auto-mode after power failure? || Have you considered to export data in a CSV (after an encryption process) to let others consult them through internet using a platform like Google? || Have you considered to add QR code in the certificate to present results or photos showing the installation of the blocks during the calibration process? || Did you use any norm (like ISO 80000) as a base to emit the certificate? || Talking about the fill in process (where you add values). Are the variables inside your software? || Talking about the software authentication process and history (not sure if it's available) did you consider user, password and a register for who and when added certain information? (e.g. who and when modified the values of temperature?) || How complicated is to modify report's structure for the final user (in case someone would want to add or remove something)? || How complicated is to modify the algorithm that evaluates or calculate data which will be displayed in the report? || Does the software verify certificate’s date as a safety measure to start the test? || Are you using a password after exporting data to Excel to protect that spreadsheet? || Regards, Angel Ochoa

For Mauricio Alberini question. Thanks Mauricio. The software's display shows what the equipment (Millitron) send in real time. Ours have got 10 nm resolution, however connecting to an equipment with higher precision the software would display the Millitron's display too. Is just a mirror.

Thanks Diego Coppa. The software was developed in Delphi 7. In latu I have developed already 6 SCADA and HMI for different calibration process, all with Delphi 7 and Delphi XE3. please refers to https://ojs.latu.org.uy/index.php/INNOTEC/article/view/464 About Python, that because I have not skills in that IDE, as well as about LabView. I'm an old romantic of development. As it says: old dog is not taught new tricks :-) Does your software read the temperature of the certificate to correct the measured values? Yes, you can include thermometers corrections in standard module and calculations will take thos into account.

Thanks Pablo. At INTI Rafael we also have a Militron, and I was able to make it show by software 0.001 µm and I would like to validate it.

>> PABLO PLEASE READ THIS COMMENT << We are expressing our doubts in this platform. The idea is to have topics to talk about and questions to answer during the "presential/virtual session" and answer them during the Workshop using Microsoft Team (14th, 15th and 16th of July 2020). As moderator, I will "interview you" using the questions generated during this week by participants in what we named: The Exchange in plenary. We will have 1 hour of exchange for Automation of laboratory processes topic (it means that in 1 hour we need to cover those 6 videos with questions and answers)

Ok. I'm sorry

Hello Pablo, thank you for the presentation, it is very interesting. Let me ask you one question, did you make the regulation of the position of the upper sensor in the comparator also automatic? Because in this comparators, the position of the upper sensor is regulated manually to match the height of the blocks, and I didn't catch if you are still doing this manually (al the beginning, when you put the blocks in the comparator) or it was also made automatic.

Excellent presentation Pablo, the manipulation of the pedal and temperature acquisition is clear to me, however I would like you to explain about how you know you are in the correct position of the block, because manually, the user feels a stop and stops manipulating the template.

Dear Pablo, good presentation. I have a question: How long did you take to develop this project?

>> IMPORTANT PABLO PLEASE READ THIS COMMENT << Due to great success of the questions and answers section during the current week. And because we are having just a few minutes during the plenary to cover all comments or questions. We decided to change the rules a little bit. We will be pleased if you could continue answering questions through the comment section (this include CENAM's platform or YouTube's comments section). Excellent job! My thanks to you and all in helping get as far as we have during this event! Regards, Angel Ochoa

>> IMPORTANT PABLO PLEASE READ THIS COMMENT << Due to great success of the questions and answers section during the current week. And because we are having just a few minutes during the plenary to cover all comments or questions. We decided to change the rules a little bit. We will be pleased if you could continue answering questions through the comment section (this include CENAM's platform or YouTube's comments section). Excellent job! My thanks to you and all in helping get as far as we have during this event! Regards, Angel Ochoa

Automation and digitization often go hand-in-hand and requires the analysis of the process stages and utilization of hardware and software tool to accomplish the most efficient outcome.

Automation and digitization often go hand-in-hand and requires the analysis of the process stages and utilization of hardware and software tool to accomplish the most efficient outcome.

Automation and digitization often go hand-in-hand and requires the analysis of the process stages and utilization of hardware and software tool to accomplish the most efficient outcome.

TO VICTOR GUEVARA >> The software generates a document with a software QuickReport which is included in Delphi's IDE. This report can be printed or saved as PDF. Moreover we included the possibility of export raw data and results to an Excel worksheet. Here in LATU, certificates are upload as PDF to the institution system.

To IRAM OCHOA >> Hope to reply all. a) Total time to software development 4 monthes, time for hardware assembly (having all parts), about 15 days. b) People involved: To software development just me, to all hardware related stuff mainly Alejandro Acquarone and me. Of course with some help of manteinance staff and a turner. c) Effective time earned: almost 100%. The automated calibration is compleately autonomous. d) Use of earned time: In the LATU's Dimensional Laboratory works only one effective technician (Alejandro) and I'm pressure responsible, so we have never much spare time. e) Final user was part of the project. But by my experience of about 7 SCADA software, a close comunication with user is mandatory. f) About learning curve, I spend a lot of time in interface design for my software so I gues (and hope) not much more that a pair of days. g) About alarms. No, when fails the software turn off. h) The software cannot move if there is no safe distance between probes, so there is no risk for equipment nor standards. In power cuts or other events the software aborts all operations. i) Export data to csv is not implemented, however the excel export implemented is not a real excel file but just tab separated data. From this to a comma separated data are just a pair of code lines. Not a problem if needed. j) Not QR code, nor pictures nor ISO 80000 reports implemented for now. k) About changes to the report, I'm afraid yed, the report template is not editable. changes shoud be done for the developer. l) The algorithm handles all parameters, variables, raw data, standards data, workflow to get results, so any change is posible but a delicated task. Of course in source code level. m) For now, certificates are audited manually befor upload to the main system. n) Excel exportation is implemented just for statistical analysis or side researches, so that's not involved in client certificate.

2nd To MAURICIO JAVIER ALBERIN >>. The Software display works parsing the output text from Millitron, which is setted to sent to print each second. So this software could show Millitron resolution no matter how many digits uses that model. It was implemented like that because we couldn't get comunication protocol of millitron from factory.

To JOSE SAMUEL RAMIREZ HERRERA >> No. Sadly the position of upper probe must be still set manually with upper knob.

To RUBEN GIL >> Thanks Ruben. In the presentation there is a part when that's explained. Before calibration there is a procedure named "Start Robot" where the actuator runs all around the themplate collecting references. A bit deeper explaination: it gets references of template limits in perpendicular directions (so here it already have both centers exactly and two random positions laterally). Finaly to get corners of sample it goes to a side limit and just turns to upper and lower limits (to corners). After that goes to the middle, cross to the other side and repeat procedure to get the other two corners. The fact is that the walk wich finally worked fine (to avoid obstacles) fine was a bit more complicated, but the idea is more or less like that.

To JOSE EDUIN CULMA CAVIEDES. Thanks José. the truth is that the version of the software without mechanical automation (Just digitalization) was developed a few years ago. on 2019 we advance to complete automation. Total development time from zero was about 4 or 5 monthes

To FRANCIS HAMILTON >> I agree. When somebody ask me about developing time in my mind appear a lot of hours not counted, discussions with coleagues, solutions to trouble that finally didn't work, the checking of tens of catalogues seeking for the best hardware, pieces that didn't fit, and different troubles discussed with a lot of people. Is much more that time in front of a PC. I do know that anyone who has got involved on projects like this can identify a lot of time besides development.

Dear Pablo, I create a library with the communication code for the millitrom, you can do everything from that library, even obey that additional digit.

The time we developed the software we hadn't got serial protocol, but only Mahr comparator user manual with simple settings of Millitron. I just found that the manual of millitron can be bought for the internet, perhaps is a good improvement. Our lower CMC are over 60 nm, so a better resolution wouldn't give us too much. But that's a nice improvement to take into account on next release. Thank you.

Dear Pablo, I can pass you the library for the Militrom, or the source code.

First of all, I would like to congratulate you and your colleagues on the work done. I have a question about the positioning of the sensing probe when looking for the 4 points of the edge. Is the positioning based on the template used? Or is it the equipment itself that detects the edges?

Increíble trabajo felicidades, me gusto mucho especialmente el software, me gustaría saber que tipo de actuador es el que utilizaron para el manejo de la plantilla y que tipo de movimientos, fuerza y como controlarlo, Saludos.

Incredible work congratulations, I especially liked the software, I would like to know what type of actuator is the one you used to manage the template and what type of movements, force and how to control it, Greetings.

Excellent work Pablo. When you enjoy making your work and the work of others easier and more efficient, the extra long hours spent is not really "work" but fun. It is wonderful feeling when you see the successful end result.

Congratulations Pablo and Alejandro for the excellent work done!! Is this an initiative a part of an digitalization plan of the laboratories? Are other authomatization process planned in the near future?

Testing comments section