Does Arduino have a future?

kowalski:
I believe it has to do with how the "product" has been developed. The software will never be industry strength (high quality) due to how it is handled by the "company". It is a lot of modified application notes and student projects.

Dear kowalski
Saying that arduino is a bunch of modified application notes etc is the usual boring rhetoric.. It's no longer like that.

Our code works in production in thousands of real projects used by real companies.

A lot of "professional" code can be declared so simply because it is closed source so nobody knows what the hell is in there but there is a LOT of crap believe me.

BTW if you see any low quality code in Arduino please lend to the community your excellent skills and submit a pull request :slight_smile:

m

@Massimo Banzi

if you see any low quality code in Arduino please lend to the community your excellent skills

I approach contributing to Arduino in a very different way. For little over three years I have developed an object-oriented framework, Cosa, that replaces the Arduino core. I have used my experience from teaching object-oriented design and operating systems, and many years of large scale embedded systems development to provide examples on how to get to the next level.

Cosa is a framework where the classes are designed to work together and with advanced variant handling. There are interfaces for both libraries and device driver developers. It is high performance with X5-X10 compared to the basic Arduino/Wiring functions. And provides many of the mechanisms that are missing in the Arduino core.

The Arduino IDE and core are very much entry level and very good at that. There is a need for support for larger projects, development processes and more advanced software concepts - the next level. It is not possible to retro fit a better architecture (unfortunately).

Please feel free to contribute to Cosa :slight_smile:

Merry Christmas and Happy New Year!

Very pleased to see Arduino SRL (.org) and Arduino LLC (.cc) starting to make amends.

Website, Forum, Product Catalogs, GitHub repos and IDE merges will be good.

Will be interesting to see how they handle board manufacturing and new partner alliances (STM).

I'm glad they are now working together. I hope the arduino srl will not use any low-quality regulator that goes into thermal shutdown when 12V is connected for more than a few minutes. Almost ruined a whole project. A dozen devices behaved strangely.

liudr:
I'm glad they are now working together. I hope the arduino srl will not use any low-quality regulator that goes into thermal shutdown when 12V is connected for more than a few minutes. Almost ruined a whole project. A dozen devices behaved strangely.

We might as well stop using linear regulator entirely. Switch-mode controllers with built-in switches and synchronous rectifier are fairly cheap now, and can increase both board efficiency and maximum current available at the same time.

I have cooked up a design that is largely identical to Uno but had three changes, one of them being replacing both AMS1117-5.0 and LP2980-3.3 linear regulators with TPS562200. That chip comes in the same SOT-23-5 package as LP2980-3.3, but it can push out 2A maximum, use small magnetics at the size of 5650 or smaller, and have a minimum power efficiency of 75% at zero load. If you are interested I can publish it under 3-clause BSD license. The same power delivery system is also used on my Uno-compatible STM32F103CB/STM32L152CC development board.

On my STM32F103CB board, with the CPU running idle at 72MHz, one power LED and a blinking indicator LED on, and a significant portion of onboard peripherals enabled, the board draws no more than 50mA when powered by only USB. And when powered by +12V barrel jack no component on the board get warm. So if we move the board to switch-mode converters we can cut a big chunk out of the energy consumption, and eliminate warm components almost entirely.

ARM (a.k.a Sofbank) announced new Cortex-M family members based on a 32-bit implementation of the ARMv8-M architecture.

Cortex-M23 which inhabits the M0, M+ space with good efficiency improvements

Cortex-M33 which seems a good upgrade to both M3 and M4 processors.

No details on clock speeds.

Atmel (a.k.a. MicroChips) and STMElectronics have both signed on as partners for the processors.
It will be interesting to see if Atmel releases MicroControllers suitable for new Arduinos. (e.g. maybe a SAMD23, SAM33X or SAM4X variant). With Arduino.org using a Cortex-M4 based STM32F469 microcontroller from STMElectronics, who knows which direction the folks will go for future boards. (we've had some toe dipping with Intel and TI by the Arduino folks... Now we have STM)

Announcements hinted at high performance variants to come (think Cortex-M73 or whatever they decide to name it) that will reside in the current Cortex-M7 space.

Arduino.org was at the "ARM TechCon" tradeshow... It was sort-of amusing to see "a parent" of Arduino present for the first time at a show where multiple other vendors have been showing "Arduino form-factor" boards for years...

ARM's announcement: Cortex-M23 and Cortex-M33 - Security foundation for billions of devices - Architectures and Processors blog - Arm Community blogs - Arm Community
They seem to be selling it primarily from an "enhanced security" point of view.
I guess it won't be clear how it stands in the power-use range until there are actual implementations (it's interesting that there are already "respected" vendors offering CM4 chips for their "low power" range, rather than going to M0/M0+.)
It'll be nice to have "divide" and some of the other instructions back...

I saw another indication of the cooperation between the two Arduino entities today. I went to arduino.org to see if I could purchase the Star Otto yet. When I select purchase and went to the Arduino store, I saw that I was now in arduino.cc land.. their store. Interesting.

Resurrecting this thread in light of the Arduino AG acquisition that was finalized today. Get to work guys.

Response to this Blog announcement.

I feel that AVR is falling behind ARM...
And now that the esp has come around it seems even further behind

AVR is falling behind ARM.

Well, that's a certainty. AVR is essentially a "legacy" architecture, while ARM is still evolving. In theory, it doesn't matter as long as the AVR is "enough" for a particular application.
I guess the big "danger" is that "CircuitPython" will get fast enough and cheap enough to implement the things people want to do, while being much "friendlier" than Arduino. So far, I haven't decided whether I'm UN-impressed by Python on a 256k 48MHz ARM with an extra 2Mb flash being "almost enough" for people to use, or stunned that those boards are the same price as the 32k AVR Arduinos.
(It wasn't that long ago that there were a bunch of Embedded Java computers that were going to take over the areas where Arduinos and WeMos board now prevail. You haven't hear much recently about SunSpots, Jini, or Javelin Stamps, have you?)

Since the OP was written in 2014, and worried that Tre would replace AVR Arduini, the Tre has disappeared and AVR Arduini continue, despite several attempts to wreck the Arduino company.

CircuitPython™ is a bit of a joke I think, MicroPython already seems to be killed with multiple incompatible forks.

Far from not having a future, I am wondering what it would take to kill AVR Arduino! Even if Microchip dropped the ATMega (which is very unlikely), there are clone AVR devices. If Arduino AG folded, there are probably enough interested parties who would pick up development of the IDE etc.

It looks like the Arduino Uno will literally be with us forever now.

Arduino may have many faults, but it is very versatile. The Arduino TRE will always be an arduino board, with all Arduino's pros and cons, and with the Arduino IDE. Arduino should update many things in the boards, but given that it was born to teach, it is very good.
This is what I think.

I hate to diverge any from this wonderful conversation... And yet, here I am wondering, what happened to

boards like the Esplora and Arduino Robot? Those were great, and with some improvement could be

fantastic. Built-in sensors and controls allow more flexibility in space management, and the potential uses

are far beyond those of a standard starter kit. I would love to see more boards like those from Arduino.

The other thing I wonder about is the standard price point. Now that Arduino is up and a few years running,

why are the boards still so darn expensive? There are clones of the Uno rolling out of China for less than

half the price of authentic ones, with no loss to quality or service whatsoever. I think that Arduino may have

to lower the cost of their products if they want to continue to be successful.

Lastly, I kind of want to see a version of a Uno or a Mega with more memory and the ability to run

multiple functions simultaneously, but without any super-fancy hard to learn code. Maybe a new library, but

nothing like a Raspberry Pi or ARM cpus... just good ol' Arduino. That sounds like a good future to me.

  • D.E.

There are clones of the Uno rolling out of China for less than half the price of authentic ones, with no loss to quality or service whatsoever.

Pleas note that China inc. does not pay for the forum and other costs the Arduino team make.

China inc. does not pay for the forum and other costs the Arduino team make.

Forum and other costs like: binary IDE repositories, github source repositories, IDE development, Core development, library development, website and documentation, advertising, evangelism, sales (in the sense of working to get Digikey to carry Arduino boards)...

Reducing UNO price to $20 would be nice. I came to know Arduino UNO as a $30 board and almost no alternatives elsewhere. That has changed a lot. The technology involved in making the board is well-established and the board can be cheaply made.

Just noticed that the official price is $22 now. Not bad.

I wonder if Arduino should change business model a bit. I know the open-source business model is very new and appealing. But it is also prone to issues such as copycats and dumping. What if Arduino LLC decides to lower board price to $10? With the official appeal, they can get rid of lots of competitions that are no more than a fab house with a couple of ebay accounts. Arduino LLC can still make a little bit of money but they will need to get the rest of the money by annual donation campaigns like wiki or others. I'd be willing to pay my contribution if I know that board quality is good and cost is low, even if I'm not buying their boards every year.

Giving a donation and trying to find the cheapest board to buy are two totally different things that I believe many people can do both.

While there are several more powerful products like BeagleBone an Pi etc, the beauty and success of Arduino IS its simplicity. If I need a bit more memory or speed for a particular task, I use an ESP-8266. Just those two choices cater for 99% of my IOT projects.

For more complex projects, e.g. a very comprehensive weather station, I've found multiple Pro-Mini / ESP nodes have often been a better solution, that's also easier to maintain. For the 1% that demand more grunt etc a CubieTruck or Pi has been more than enough. Beyond that I just use a small industrial PC with Linux, Android or Windows.

For me, the real magic of Ardunio is its totally open source (hardware and software). I prototype with standard Arduino hardware and when its time to produce a quantity (10+), its so easy find an existing (public domain) Eagle file that's closest to my needs, add on any extra logic etc, and get some boards made. All the components are readily available, it a piece of cake!

I wouldn't even consider using some of the so called 'better' boards for a more demanding project, the few people I know who went down that road regretted it due to the lack of information, code examples, and special components etc, made it a nightmare. A small Linux / Windows single board computer is quicker and more reliable etc.

So, yes, I see a long future ahead for Arduino, its simplicity making it perfect for the explosion of IOT devices.

Just my 2 cents worth on the topic.

Cheers!

Paul

I don't love the sound of that either. I wouldn't like to see Arduino move in to the micro-pc space if it means leaving the mcu space behind. I have Raspberri Pis and Orange Pi Zeros and they are great where they belong, but I would much rather see Arduino move to better MCUs and stay where they have done so well. I don't think that sticking with the 8bit chips is a good idea, but I would love to see some good quality boards build around A LOT of new chips. They're packing everything in to some of these new MCUs. I'm waiting for them to start integrating e-paper on the dye or some such.