When you have to abandon a project? When is it worth pushing on? The sunken cost fallacy at work.

When you have to abandon a project?
When is it worth pushing on?
The sunken cost fallacy at work.

Over the past 12 months I tried to turn a pile of random parts into a working 4 axis CNC machine.

You have seen me deal with a ton of wiring, control concepts, mechanical arrangement, bolting stuff together and mounting issues. But also Software limitations and complications.

When I started this Blog series my hope was to finally turn a pile of junk into a useful pile of junk.
At that time I was of the assumption that my workflow would greatly benefit from having a 4 axis CNC. That was the theory.
But in reality in the past year I have gathered much more experience with combinations of Laser cutting, 3d printing, PCBs for structural parts, standard pre-made enclosures and Sheet metal fabrication as well as CNC services.

And I concluded that 90% of my workflow for any given project so far, does not benefit from such a machine. Furthermore generating suitable Gcode and operating the machine to a particular standard of surface finish and speed is very difficult and a whole new hobby or profession in itself.

I finished a few client projects in the past year that were dependent on accurately manufactured parts and for these I tried out CNC services and learned to design for bent sheet metal fabrication.
If I conservatively calculate my hourly cost of setting up a CNC job, material and running my own machine assuming no failures and an already perfectly calibrated machine + known speeds and feeds for each material. I end up far above the prices local services quoted.

But I was already to deep in the trenches to back up.
From operating my tiny 3 axis CNC, I knew how hard it is to get the right speeds and feed-rates for a project and it would almost never work on the first try.

But realizing that I’d pay more for the raw material than for getting a finalized pieced shipped to my door has put the nail in the coffin.

I learned a lot about wiring, enclosure adaptation, setting up high power Stepper drivers and I have built a self contained power /control unit, ready to be put into a rack mount enclosure.
It can run a wide array of Stepper and BLDC motors while accepting normal reprap-firmware Gcode.
So the progress of the project is not lost, it just turns out this part of the project can easily pivot to a new function. Whenever I want to try out an old salvaged motor for function or just need to control motors for a temporary setup (like during filming) i have the almost perfect solution for that handy.

The sunken cost fallacy drive me towards a wall.

Let’s pretend that I have finished the device mechanically and electronically.

What now? How do I generate Gcode for this unique design.
I’m not a skilled programmer, writing a Gcode converter or a custom setup solution is far beyound my skillset.
I looked at various solutions but ended up either in the corner of this works for 3 axis CNC, but not for my weird 4th axis, rendering it almost pointless or degrading it to a position changer in between cnc operations.

Or the documentation was so confusing to me that I just could not figure out how to implement what was shown.
And with these roadblocks ahead I just don´t see a point in going forward, when my main focus has to be getting projects done that I depend on.
So whatever helps me get the project done is the way to go, and this CNC is actually getting in the way of my goals as it takes up a huge spot in my tiny workshop. I will let it sit as is in a better suited corner, so maybe i have a need for some oft he parts I already fabricated, I can utilize them.
But for now I need to learn my lesson and focus on the fabrication techniques that give me a lot better outcomes for much less hassle and in case I need metal fabrication, a quick search for material prices will usually lead me on the road to decide for service, a different method of fabrication, another material or maybe ,but just maybe , at reviving this CNC project.

So what does this all have to do with electronics, prototyping etc.?

It boils down to a hard to understand principle.
The sunken cost phalacy.

Especially in electronics, we often strive for ideal solutions, elegant designs, or cost effective compromises.

Very often I see myself obsessing about a tiny aspect of a device, getting really excited to have found a solution that would perform a tiny bit better or more efficient, only to later realize that it drives up the parts cost massively, but does not really matter in the grand scheme of things.

It is similar to feature creep, where you just add more and more features aspects or subsystems to a device until the main point of the project goes on the back seat. You lost focus.

When you loose focus on what really matters, it is very easy to refocus on the wrong parts.
What is the actual goal of your project? Does it even have to be a dedicated device or could a python script on your pc do the same thing?

When you have invested a lot into a project, money, parts, but also time, people are very inclined to just push on because the goal could be near and they have already come along way.
What is true in hiking must not be true in electronics.

When I wander for 4 hours up a mountain and I’m resting exhausted just below the tip of the mountain, it is easy to see how just walking on for 2 more hours will give you the reward you worked so hard for. And also you are very aware that if you give up now, you will have to walk all the way down the mountain again. Without any triumph, without victory, just more cost in time and pride.
So why not push on so you can return with triumph, pride and great photos from the top?

If you turn back you loose double. The time an power it took to go up and to go back down.
So you push on wards and overcome the hurdles no matter how big to reach the summit. And hopefully not die trying to hard.

But electronics design is not hiking.

When you are deep into a project ,spent hours in ecad, bought sample components to verify pinouts suitability etc, wrote multiple firmwares that each test a specific aspect or component of your design, standing in front of an unclimbable wall is a bad feeling. Even more if that wall is not a design aspect or a component choice, but just lost focus on what actually mattered for the project.
It is like having conquered a steep trail up a mountain, and you do not want to loose twice by giving up!
So many hours have been put into BOM optimisation and current control, even the sourcing troubles gave you a headache, so you just don´t want to let it all go to waste!

But the crucial difference is , there is no way back down, there is no cost included with stopping.
You are not loosing twice like on a hiking adventure, you are only cutting losses before they accumulate to high. You have only lost ,if you haven´t gained anything on the way, or at least stop loosing at this very moment.

The samples you bought can be reused for new projects, the time you spent in Ecad honed your skills, the BOM optimisation is a great starting point and reference for your next project.

You have not lost.

Yet.

As you proceed in your project you are only gaining experience, skills and knowledge. You are only loosing if you don´t know when to push on and when to move on.
It is called moving on, not giving up.
It is trashcan not trashcan’t.

Your are not crawling down a slope in shame, you are taking your assets to a new mountain range where they help climbing, where the goal is achievable or actually matters, and you have only won up to this point.

I have spent 10 years in a Job, just to leave it behind and start this self employed electronics thing. Did I loose by spending 10 years on something ? No , I just moved what I have gained to a new profession. I didn’t have to walk back and start again, I took what I gained and started at the foot of a new mountain.

And it is the exact same thing when you develop a product or personal project, if it does not work out and if the focus has been lost, you have only lost if you push on to far and sink more and more capital into a drowning ship.
Knowing when to move is hard but healthy. For your mental state and your financial situation.

I could spend €400,- more on material and parts for my CNC, but what would I get even if I press on until the bitter end?
I would end up with a machine that maybe works as intended, is very hard to program, and running it would cost more than just ordering custom parts from people who are specialised in operating these machines. So the expected quality will not match professional offerings at all.
My goal was to get parts for projects, not a lifelong project.
And somewhere along the line I forgot the projects actual purpose.

So I move on to the next adventure and cut my losses, take my honed skills and utilize them in a new season of element14presents.
So have I lost, or have you lost as a reader of my cnc Blog series?
NO.

We both won the same way. We both got to experience the steep trail of building such a machine, we honed our skills we gathered information, understood principles along the way we did not have much to do with before, and now at the very end we have gained a Blog series, valuable feedback from the Element14 community, and again a huge pile of parts handy for many future projects.

Thank you very much for joining me on this non hike into CNC land.

Moving on

Clem.