Developing Hardware is… well HARD!

For the past several months we here at AmperageApps have been developing and refining a prototype known as “Miljø” (“environmental” in Norwegian) which monitors the environmental conditions in a home or office.  The system runs on a dual core 32-bit microcontroller and is chock-full of sensors for everything from CO2 level, UV light, sound, air quality and motion.  The project began as a prototype for monitoring construction sites for environmental compliance and has expanded to become an industrial device suitable for construction sites, office buildings and manufacturing facilities.

But let us not get ahead of ourselves here; it is one thing to cover the office walls (and a few windows) in post-it-notes and scribbled schematics, but quite another thing to see it sitting on the desk and working.


This blog post should be entitled
“When software developers try hardware”.


We must start at the beginning and that seems to always start with “here hold my beer” …  Oh wait, no, sorry that was another story. This project started with an idea of how we could improve worker safety and build a platform for analytics and data collection revolving around worker safety.  We have been working for some time on our AMPDrill+ Offshore Drilling and Analytics package and with the recent collapse of the oil market we thought it prudent to turn our attention to other industries.

The initial phases of development we went through:

  • Idea
  • Planning & Research
  • Budget
  • Component selection and functional design

We set about gathering parts and then we hand-soldered and assembled the first two prototypes, which allowed us to discover what we had missed or forgotten.  Its funny how things seem so obvious after you realize you forgot them!

Next, we worked on making changes and revisions based on testing and feedback.  And after that it was time to go shopping:

  • BOM (Bill of Materials) determined
  • Circuit board designed and ordered from fabrication shop
  • Order parts (and wait…….)

Now along comes our new reality of the corona virus and everything comes to a grinding halt overnight.

At this stage we had entered the corona virus lockdown along with the rest of the world and we knew we would be waiting a long time for parts from our suppliers.  This worked in our favor as we turned to software development.

We built an entire backend stack that we call AMPVision Data Visualization Service and developed a dashboard architecture.  After this we developed the microcontroller firmware and sorted out the sensor readings. We learned all we could about calibration and testing of sensors.



3d Print the first case design (12 hours) – did not fit the board – umm, Houston, we have a problem.
 
3d Print the second case design (15 hours) – it fit and worked on the first try!”


At this point we had 10 shiny new circuit boards on the bench that arrived very quickly after the order was submitted. Over the next 3 months all the small part we had ordered began to trickle in.  

Finally, everything was now in place and we set about hand-assembling the first unit that we called Prototype #1.

Circuit board made by AmperageApps

Prototype #1:  We hand-soldered and assembled the unit in a late-night coffee-fueled rush and with much excitement and anticipation we applied power! That is when we experienced the cursed “magic smoke” of doom. 

As we would discover later there was an incorrect trace on the circuit board that caused a short circuit and that was the end of prototype #1.

RIP Protype #1 AKA Big Fail!

Prototype #2: AKA Let us try that again! Lesson learned; test each stage of the board design iteratively and never assemble the entire system in one go.  We learned to test the board and the power supply first, then assemble the passive components and test again.  Then finally install the microcontroller and other “sensitive” bits.

Hurray It worked!

I will jump ahead a bit and say that we have learned a lot and come very far in the journey to arrive at this point in the development of Miljø. The project is and has been a technology incubator for us in that we now have a fully functional IoT stack and customizable firmware available that allows us to securely read from multiple sensors. 

The bigger challenge now is to decide what we will do with the Miljø concept. Should we proceed with it as a commercial device or should it remain a launching ground for other projects and designs?

The lessons learned and the education we have gained is allowing us to begin on several other designs and projects that are now taking center stage on the workbench. 

We are working with other developers who are interested in using our IoT stack for their product and the hardware development experience is allowing us to create some interesting designs so stay-tuned for details on those coming soon!

We Create Solutions and make Technology, Sensors and Data easy to Use and Understand

We are Problem Solvers, Consultants, and Champions of Cool Technologies

Leave a Comment