What stops your Arduino project?

I feel very comfortable with the software, it has no mysteries for me. In general, adding sensors and transducers that already come in modules with I2C or SPI communication is also easy.

Where I always get stuck is how to power the projects. How to calculate the necessary power, what options are there to make it portable. Also how to reduce EMI interference and how to protect the circuits.

I run into problems when sensors or peripherals that I want to use don't have libraries available for the Arduino IDE and it is necessary to port a library or develop one from scratch.  I also have problems when there is no board library available for a specific board implementation and it is necessary to hack one from another board using the same MCU.

The inverse problem also exists, where there are too many libraries available for the same board or peripheral (with different features).  If multiple of these libraries are installed and definitions are reused it becomes a real pain as the compiler gets confused.  I encountered this a lot with the many custom library versions for the RP2040. 

An Arduino critic once said...

"Its fun to have tonnes of arduinos, even though you can run only a single program at a time, the other day i was playing with nano 33 ble, trying to run mbed rtos::dual-tasking, it miserably failed for nothing although the core is meant to run on mbed platform. Libraries like ArduinoBLE have basic functionality and are painstakingly bitbanged as per the use case. Hence in my opinion the stack and software repository needs to be updated frequently."

From some of the discussions on the Arduino forum that crop up, I think videos or other info teaching users how to structure their code design could help. Sometimes beginners use a hundred lines of code when they could have collapsed it all into a 5-line 'for' loop for instance, or creating functions. Also, how to troubleshoot code - from blinking an LED or two for status, to using a multimeter or 'scope to check activity on pin(s), and using the console for debug print statements.

Also, how to test algorithms using a PC, Pi etc., before placing that code into the Arduino IDE. Using #ifdefs, etc.

I think a lot of people could concentrate more on the end application if they can get through the coding basics, before working with the Arduino libraries.

Also, some example code structures/patterns, for instance how to accept user input, maybe debounce it, and also do some processing in the background, how that could be implemented with timers or interrupts or any other service provided by the Arduino system to achieve such real-time tasks.

I find the Arduino libraries helpful for most of my simple projects.  Unfortunately, there is much poor Arduino code on the internet and care must be taken to find a good source.  So, a list of reputable sources would be helpful.  There are many processors and boards now so care must also be taken to understand that some code is platform dependent.  My own opinion is that it would be better for beginners to start out with newer 32-bit chips (e.g. Nano 33 IOT) than the older 8-bit AVR models as they are priced well and are more powerful.  I often use different sources of power, motors, WiFi, and sound in my projects and these subjects aren't always covered in the basics.  I believe that interrupts and coding without the delay() function should be introduced early.

When I got started with Arduino several years back I hadn't written any code in more than 40 years.  And that was FORTRAN on a mainframe.  I started with the SparkFun "Inventor's Kit" and it had me up and running with the basics in one weekend.  I then worked through the examples in the Arduino IDE and made sure I understood the functions, structure, variables, etc. that are in the Arduino Reference documentation.  Thinking back, the advantage was that the kit had a well-written guide and all the required hardware.

My hurdle is coding. I started in electronics and my career progressed to system support. Learning to code was something that help with system administration. I have invested in learning how to code better but it does not come naturally to me. I developed a whole security system in Perl but find it difficult today to read the code. It is hard for me to believe I created the system. Individuals who have worked on the system have mentioned how easy it was to understand and make changes to the code. That befuddles me because I wrote it and don't understand it anymore.

Usually some endless loop. Sometimes sudden overvoltage ...

- Gough

I just fired up Arduino and it printed this:

void setup() {
  // put your setup code here, to run once:

}

void loop() {
  // put your main code here, to run repeatedly:

}

Which is ok, there's some help, but maybe needs a structure to help. For instance (this is just a crude example I just typed now, but there must be better examples out there):

/*********************************************
 * My Application Name
 * revision 1.0 - date - name - description
 * 
 *********************************************/

// Includes
#include <Wire.h>

// **** Definitions *****
// general
#define LED_PIN 0x01
// I2C addresses

// other GPIO registers

// convenient values, states

//******* typedef structs ********


// ******* consts ****************


// ****** global variables *******

// ****** function prototypes ****

// ****** setup function runs at startup *****
void setup() {

}

// ****** loop runs forever after setup *****
void loop() {

}


// ****** other functions *******************

And then when people start with a neat template with comments etc., then that kind of makes them want to continue to write neat code for the remainder, and just seeing the different sections makes one realize to read up on those sections because they are probably there for a reason! : )

A simple debugger. Step, breakpoint and watching variables is a good start. A debugger is available for some boards now as a separate kit.' I've missed IDE debugger integration since day 1. 

Low power designs. Most Arduino Classic designs run the controller in high power mode, and converting those to low power is not straightforward. Improved with the MKRxxxx family.

Awesome question. I do lots of arduino projects so the issues below aren't usually big enough to be show-stoppers for me.

For me the biggest issue is that arduinos aren't always the best controller for the job, even though they now have a wide variety.

The next biggest issue is programming in C/C++ - just about my least favorite language. (LISP and APL are worse)

The next biggest issue is everything except the arduino - interfacing, building packaging, mechanics.

The next biggest issue is buying all the bits - displays, power supplies, motors, interface cards, enclosures etc. that can far exceed the cost of the arduino.

And lastly finding time is always an issue.