I hope the topic is not off-topic here. I am looking for a simple prototype inventory/management system. What I need is a way how to better organize the process of testing prototypes and final products (we do small batches of 50-100 final products).
Boards arrive from a manufacturer - they can be fully assembled or may need some work
I label every board with serial number
Now we perform a bunch of tests on every board to see if every component is working. We also upload proper programs into MCUs and others. Some boards may also need some finishing tasks (i.e. soldering connectors)
We distribute the boards to different people - mechanical person, design person, install them for our customersā¦
So basically I need to have more overview about steps 3 and 4. I would like to see where is every board, which tests did it pass, what software is uploaded into it, if it is considered a prototype or finished product, etcā¦
The most important features are:
Custom defined fields (for tests and finishing tasks needed)
Tags or any means of grouping and filtering (by project, by state, by tests)
Comments for every item to be able to have a history for every item
So far Iāve been using spreadsheet and it has been functional but it is too cumbersome. So I would like to use something more suited for the task. Iāve seen https://github.com/inventree/InvenTree in these forums and even it looks like a great tool I am not sure if it can do what I need. It seems to focus more on the manufacturing processā¦ Another option would be to use Redmine as we are already using it as project tracker but so far I did not find any plugin that would be much better than spreadsheet. So the best would be a wiki which is not good for grouping/filteringā¦
I would also prefer Free/Opensource alternative but I do not mind paying a few bucksā¦
We also do our products in small batches of 50-100 products, and also have them manufactured by contract manufacturer.
I think you are asking for much more professional solutions than we use but few words that may be you will find being helpful.
I have written (Builder) our "production programā. It has the current version of source codes for each product and loads it into products keeping the track of numbering products.
Also the testing procedures for devices (unfortunatelly only for some of them) are included in that program and each test generates the report file (containg the device number) which can be backuped if needed. But more important in our situation (I am 200km from our production) is that if at production they canāt solve the problem they can send me te report to help them.
It would be not a big problem to add to that program some kind of keeping the history of all actions done but as always there are more urgent jobs to be done.
Currently we are at stage of some modification of our production process. In future that program will load not the destination software into devices but the testing software. Then the other program using standard device interface (having not access to in circuit programming) will be doing:
the testing,
the destination software loading.
That idea we had many years ago, but as always we had no time to test if we are able (from device interior) to modify the bootloader section and then lock-it.
Thanks for your answer. In my opinion I am not looking for too much professional tool, quite the opposite. All I am using niw is spreadsheet with filters and collumns like date, serial number, lcd tested, mcu programmed, mcu sw version, all buttons ok, Raspberry Pi working, Raspberry Pi serial, current place, current responsible person, ā¦
And basically I would like to have it online and with historyā¦ So I can answer questions like āI have K123 here, is it working?ā or āI need 5 prototypes to give to external designer, they does not need to be working but have to be mechanically correctā
I could program something on my own but it would be a lot of work and I think that system this simple must be already developped somewhere
I have mucho (decades) experience with this (as both an engineer, programmer and company director)ā¦
Programming what you want in Java/using SQL would be a challenge and not quick. Programming it using Access or, even better, OpenOfficeās Base, is very easy (and, free).
Check out this link (or other links) and try making oneā¦ In OpenOffice - they are fairly simple and quick (and can be as robust as neededā¦). Not to mention becoming an added āToolā to your arsenalā¦
Thanks for your answer.
I have mucho experience with Delphi and I could create such app probably as fast as in OpenOffice Base/Access. However I still think that system like this must be useful for so much people (not only electronics) that there must be something already made. And most importantly something that is already āfield testedā. The worst scenario is that I need to fiddle with my app instead of using it to solve the immediate issue.
For example in one of our product we use Raspberry Pi with ATMega MCU. We did not test the RPis as we never encountered a faulty one. However weāve got some problematic batch and like 60% of them had issues with some GPIO pins. And then we had an issue on a soldermask on the MCU which caused bad solder joints on some pins on some MCUsā¦ And that is the moment when I need answers fast (āget as much working RPi+MCU combos as possibleā) and I would not be able to tweak my system to suit the actual situationā¦
It wonāt be cheap but consider using bar codes and scanners. Manually entering data leads to errors and spreadsheet expose all the data to corruption.
I think perhaps youāre not well informed about Database-making aspects of some software out there.
Iāll focus on OpenOfficeā¦
My experience with doing this includes:
Product inventory and production management of Electoronics, Wafer (chip) assembly, Electrical connectorsā¦ labor-hours, costā¦ the list goes on and on for companies I worked for (JAE, TYCO, AMP, BenchMade Knivesā¦ Many of my JAE connector designs are in Kicadās partsā¦). My customers included HP, Intel, Tektronicsā¦). Over decadesā¦
If you walk through the tutorial/link, and build the example, youāll see it completed in 10 minutes. Along the way, youāll see other options in the selection panels and checking them out will make it clear to you. You can dial-it down to the exact product names, serial numbers, revās, build, programming, status, location, etcā¦
You can build all the fields/names you want - I just did the DB below in 4 minutes following that linkā¦ (I selected different fields from the tutorial - I could have used field names you want but didnāt botherā¦).
Itās worth the 15 minutes investmentā¦ And (after seeing previous post, Barcodes are included in the selection panelā¦)
[EDIT] My memory is oldā¦ totally forgot to mention that Building queryās to find/matchā¦etc is done via the Query building panel. Very little coding (if any) - pretty much Drag&Drop kind of stuffā¦
Thanks for the example. I am kinda aware about the database software, I did my first Access database app in the previous century and I do use Delphi to create database apps as part of my dayjob for a few years nowā¦ But my experience is that creating the app is the most simple part of the whole process. It is way more important to understand and include the ābusinessā logic into the app. And that is what I am looking here for as I suppose that I am not the first to have this issue.
P.S. The reason I prefer something like Delphi to Access/OpenOffice is that I feel more comfortable investing my time into something where I do not have the feeling that the tool is limited for a specific task. It may take up a little more time (not much) to code something like this is in Delphi and its visual designer but I have full power of complete development platform when I need it. For example would you be able to connect OpenOffice to a serial port (for example to read ATMega serial number and fill it into DB field)? But that is just may opinion and your mileage may varyā¦
[Edit] And as I was writing the answer to the other post about barcodes, I did remember one more thing. Is it possible to run OpenOffice app on the mobile phone? If I would go the route of creating my own app, it could be a great idea to have mobile app that can read the barcode and do something (i.e. set the result of the test or lookup history of the item)
Thanks for your suggestion. The only issue with barcodes is that you need a scanner to read them. But these days with mobile apps for barcode scanning it can be a viable optionā¦
Smart/MobilePhone & Tablet:
I wrote some Android Apps that use SQL database and Java. One for tying Knots and one for Astronomy are posted and I made a small fortune with the Knot app before changing the cost to Free.
Itās all very doable and can include Barcode scanning from the smartphone.
I also wrote an Open-Source desktop Document Control database app and recently removed it from the open-source site because it was very popular in France and I was getting far too many emails (offers for work/etc) in French and translation/responding became more work than I want.
Writing a Java/SQL app can be fun and Data Input&Mgmt and Query can be in an all-in-one GUI (examples below).
Having said that, my experience in this area tells me that, if willing to Pay for an app, it wonāt be long before youāre paying to have it tweaked or, will be looking for another, more suitable app. In the end, youāll invest more time and cost and, though you may get what you want, similar/less effort would benefit you/your company by taking the challenge.
Many Youtube tutorials to help and Iām sure youāre skilled enough to make quick progress.
Recommend as a starting point: Grab a sheet of paper and layout what you want the GUI to look like. It will spawn ideas and questions to ask when looking for software or designing your own.
Start with something like OppenOffice or Accessā¦etc before considering doing it for Mobile.
The mobile app would be very similar to a Java/SQL (they both use Java and sql lite).
Regarding connection to serial: Yes, through Java - donāt know about OpenOffice/Access
Heck, if you post your concept, I may feel inclined to prototype a demoā¦
Well, I would argue that the barcode sticker should also include the information as human readable to allow human intervention when (not if) the scanning tech fails. Also, after identifying (for example) that serial numbers 42227, 33154, and 77516 are the ones you want to pull it can be faster for a person digging through a pile to recognize those 3 numbers using their natural scanners (eyes), than having to scan and then read the result from a technological scanner. Use the barcode scanner to quickly identify and pull up the record on the computer so the status can be verified and/or modified. But the human readable part should be a viable backup.
You can read Barcodes directly from a smartphone (I do it often at HomeDepot) and use them to search for product infoā¦ a very common event.
Below is a Database re-fresher for me (done in OpenOffice) - killing time.
Itās crude and took about 10 minutes. A query panel can be separate or on the main GUI panelā¦ The Barcode/QR codes are dropped-in images (though they could be scanned into the panel)ā¦
Iāll let it rest here but will answer questionsā¦
(I know the guy behind it [cc: @jwr], heās really nice and super responsive; for you the free tier will most likely suffice, if you see value supporting the product or need additional more important features, he also has a couple of paid tiers)
Hey @Kedarius , Iām the author of Inventree and Iām glad that it has caught your attention!
Your requirement for keeping track of tests you have performed, firmware versions, etc, is already a core component implemented in InvenTree.
I work in a small electronics engineering company, and we use InvenTree integrated with our automated test procedures. You can keep track of individually serialised parts, and record test results against them. Examples (but not limited to) that I use:
firmware version
device settings
calibration data
acceptance test results
When a test is added (either via website or API) it automatically records when the test was performed and who performed it.
Then, you can print out a test report (which is completely customisable by HTML or LaTeX templates) for that item. We use this as a final CoC before shipping.
It can also print out custom labels too.
Thereās a python interface to the API meaning you can extend the functionality as much as you want, and incorporate it into any test fixtures you desire.
Plus InvenTree also does full BOM management, stock control, etc.
I think it is certainly fit for purpose, let me know if thereās anything specific youād like to know.
(Note: Iāve put a lot of work into this but the documentation is sadly lacking! Definitely need some assistance with docs but donāt have the bandwidth)
+1 for Inventree. Iāve contributed to the project a tiny bit, and I think itās VERY well suited to a small shop or individual - both are ways in which I use it. Thanks, SG.
Sorry for the late reply, I had to finish some urgent workā¦
Iāve looked into OpenBOM and partbox.io (mostly partbox). I tried partbox demo and it seems like a good tool but it ends where I need to start. Basically it has ābuildsā where you create some number of units āunitsā. Now I would need to manage the units created.
We focus mainly on software development so we prefer to offload as much of HW assembly to a manufacturer as possible.
I will look into Inventree and I will report back
I played with inventree last night and I have to say it is a really good piece of software. I have a few questions and feature requests. Should I post them here or on the github?
Have you tried Partkeepr? Itās open source but I think it could do with some more help getting it moving again if anyone is interested in helping to develop it. We use it and itās very good. There are some features that would improve it, but itās good for tracking lots of componets or products, BOMs, Metaparts (e.g. resistors where you want to group them by spec rather than manufacturer).
