DEV Community: Miguel Tomás

live data acquisition device (LDAD) for estimating concrete maturity

Miguel Tomás — Wed, 18 Aug 2021 14:57:27 +0000

Is finally here, the #PCB i ordered for the #concrete #maturity sensor project.
With it. The final version of the #prototype is ready to deploy!

openSource #openData #IoT

The sight seeing gift of a trader

Miguel Tomás — Tue, 03 Aug 2021 16:37:12 +0000

The philosopher's stone, the holy grail , in stock market predictioners is to find the mathematical formulae that can predict the future. That said, no trader or business ever in history was able to achieve that with only historic data . To achieve sight seeing of future market valuations is needed something completely different . Is needed a dataset with expected production outputs for the near future . ...
....that said,
The non linear regression below is not that good. One can achieve better results with a simple deep neural network architecture rather than the one employed here. The only obstacle is for one to find the optimal number of hidden layers and optimal number of neurons on the network. One that is able to output a regression result of near 100% accuracy .
Enjoy !

New open-source challenge: home automation

Miguel Tomás — Fri, 09 Jul 2021 10:16:44 +0000

This is a really #cheap robot mop cleaner. And is the best device to benchmark what is a good ,functional home product. I've been testing it this morning, and to my surprise it does what is suppose to do : wet and clean the floor. However, don't assume because it's a robot it will do all things for you. It doesn't. To get the best cleaning results, the mop needs to be literally water saturated to a point of dripping. Also, another non documented trick is, the floor cleaning agent / detergent. I have mine on a spray bottle, and gave the mop a good spray all around before pushing the start button.
Since this robot cleaning does not feature a path finder / trace , the best course of action is to limit the cleaning areas and divide spaces into small 5 or 6 m2 space areas. Finally under sofa and under the bed is now properly cleaned with this cheap $30 cleaning robot.
As usual, here's the challenge:
Products like these are the best to give a boost when prototyping and doing product development. During upcoming weeks ill be adding the functionality of knowing where the robot is in the room , "spacial awareness", and where it has passed and mopped the floor. Ill be using an arduino nano sense ble to achive this. At the end all will be accessible on a open-source Android app and also on the edge computing home server I'm already building (open data and source). Total cost: around 50€.This is a really #cheap robot mop cleaner. And is the best device to benchmark what is a good, functional home product. I've been testing it this morning, and to my surprise, it does what is supposed to do: wet and clean the floor. However, don't assume because it's a robot it will do all things for you. It doesn't. To get the best cleaning results, the mop needs to be literally water-saturated to a point of dripping. Also, another non-documented trick is the floor cleaning agent/detergent. I have mine on a spray bottle and gave the mop a good spray all-around before pushing the start button.
Since this robot cleaning does not feature a pathfinder/trace, the best course of action is to limit the cleaning areas and divide spaces into small 5 or 6 m2 space areas. Finally under the sofa and under the bed is now properly cleaned with this cheap $30 cleaning robot.
As usual, here's the challenge:
Products like these are the best to give a boost when prototyping and doing product development. During upcoming weeks ill be adding the functionality of knowing where the robot is in the room, "spatial awareness", and where it has passed and mopped the floor. I'll be using an Arduino nano sense BLE to achieve this. In the end, all will be accessible on an open-source Android app and also on the edge computing home server I'm already building (open data and source). Total cost: around 50€.

o modo mais seguro de gerir passwords em 2021

Miguel Tomás — Mon, 05 Jul 2021 15:45:06 +0000

o modo mais seguro de gerir as passwords, é deste modo:
1- utilizar o programa acima para gerar passwords com o maximo numero de caracteres permitido pela aplicacao ou webpage. por exemplo o facebook permite passwords com mais de 500 char . o Google passwords (base64) com 100 char, etc...
2- guardar num ficheiro TXT as passwords geradas
3- copiar o ficheiro TXT para o tablet e smartphone

4- sempre que pedir um password ir buscar ao ficheiro TXT
5- sempre que se esquecerem da password , optem por aceder a conta utilizando a opcao de envio de um SMS para o numero de teloemovel. Quando pedir para alterar a pasword, igonar (se der)

rearranging code on an Android project to accommodate new Apps

Miguel Tomás — Mon, 05 Jul 2021 13:45:23 +0000

This is a video of me rearranging code on an Android project to accommodate new Apps I'll be coding in the next few weeks...

https://youtu.be/aaLoLkCmD3s

Why you Must "Buy me a Coffee" to Someone and Anyone. An essay about Explainable AI on Open Data and Open source Environments

Miguel Tomás — Fri, 02 Jul 2021 07:51:35 +0000

In the midst of the G.D.P.R., there's something about authoring, ownership, and fairness of working online, more important than anything else, and is at the same time the foundation for privacy...

Recently I attended an online event. The main theme was about the dangers of artificial intelligence, in particular on promoting fairness, not only in institutions and government affairs but also at the workplace on a common business and enterprise. The speaker went on for almost five hours, explaining what is being promoted in the European Commission towards the challenges I mentioned in this article: The need for algorithms, in particular, artificial intelligence and machine learning algorithms, of being publicly explainable away from the “black box“ way of doing business, that is to say, away from “industrial secrecy” of the recipe that lies inside the AI model. However, at the end of the event, I felt the speaker didn’t address the theme properly and failed to address the underlying concern before anything else: the definition of work and its new meanings, one that includes all outcomes of all these technologies, nowadays everywhere and around us.

By all outcomes, I’m referring to the simplest and smallest task a citizen online, a.k.a. “user”, can take and do: writing a short message and publishing it online, publicly accessible to anyone. Of course, the message from an author can assume many forms, for instance, be a “comment” in a “post” on “social media”. The author can also make the message larger, bigger, with more words and sentences. Can include an image, a photo, or even attach a video to it. And the process of creating authentic content does not stop at the end of the message or “comment“ about being published. The author can also add a #tag or even add a web address link to another page, which the author finds useful for some unforeseeable reason to some but makes sense to others.

All of it involves a task: and the simplest and smallest one is to take some time to type each individual letter and emoji and at the end push the button to “send“.

I’m making a pause here to include the definition of Task I’m using here. One which includes definitions of Work, including the economic definition of Labor and Human Capital. And, as anyone may argue rightfully, attached to it, is as well, Time, in particular the amount of time the author gave or even dedicated for publishing online.

But why is so vital to think at this level, when referring to the need for fairness and explainability of tasks, whether is done by human capital-labor, whether is executed by an autonomous, automatic system, or machine?

Well, may not be evident on the spot when looking at a simple tweet on Twitter. One that has a maximum of 280 characters. Unless one dedicates enough time to think (and meditate) about it. While on it, the author of that particular short tweet added another one, one which included a high definition photograph. If you’re a professional photographer you’re going to understand what I’m about to say more easily. The amount of dedication expressed as a function of both time and work-task activities cannot, and should not be neglected nor discarded when arguing such matters, not even at the higher level where the European commission usually is. It asks the author, as an online user, to study and learn the trades of photography, the expenses needed for that photo to happen, including pieces of equipment, preparation tasks, etc. The same, when dealing not with a photograph but instead while filming a video for later posting on the user’s YouTube channel for everyone to see. And finally, can be generalized to any other task, by any other user even those not working on media or marketing. Furthermore, nowadays, remote work is no longer a trend of a few but something to take seriously for its many advantages, far beyond costs and profits, and technology is demanding as well too on the data being asked to be collected and stored on those big databases, more importantly, the constant update of each data field on such databases.

So the European Commission decided a few years back the need for such work activities to have a financial return to its owners. Back then, the concern was addressed in terms of “Big Databases“ companies such as Apple, Google, Facebook, … were amassing and turning into large amounts of profit while not being authors of the being data collected. The reader at this point may be noodling in agreement on what I’ve been writing while saying, “indeed I also want a piece of that pie to fall into my personal bank account”. The question that failed an answer to this day, is, how much should the author ask for that particular message, for that particular tweet, since our monetary system only considers fractions of the unitary coin down to one cent.

The answer to that question has been demoting further advances on this matter. Simply because the actual value of a message like a simple tweet is much lower than one cent. Unless you, the author, consider the possibility of being paid not in the traditional FIAT currency but instead getting paid in a digital “coin“. Facebook attempted this a few years back when announced the liracoin only to be forced to put the breaks on it by the US congress. The major concern was, and still is, it may become more “strategic“ than the US dollar itself while at the hands and control of a few. Not a good setup for a democratic and free society anyone wants. Since then, the FED and more recently the ECB arrived at this same conclusion, also due to many other factors and needs out of scope today on this article, the need of having more smaller fractions of a coin to a point of a single tweet can be quantified and measured as paid work. While that is being defined and decided, the reader may not be paid for a single tweet today, but how about an online video on YouTube publicly accessible to everyone to see? Google (hats off here to google) came to the conclusion that it was in their best interest (for many reasons, which includes courts and heavy fines) to pay for content YouTubers were making available on the platform.

However, is not enough, demanding big corporations to pay for data they collect. Is also needed, for you to learn how to be paid as ask to get paid for this type of work-tasks and activities. For instance, give an hour of the reader’s time during the week, to explain a subject to someone or a group of people. What about the cases that revolve around closely related? Friends, Family, and acquaintances? Does make it an official business or can be seen differently, something more close to charity or a donation? My answer is the latter, and all those involved should be made aware of that, in particular, the advantages to all, attendees and speaker (for cases of an online event) and even on cases of really small daily tasks such as helping, on the spot, an elderly having difficulties with groceries or picking up a piece of harmless garbage on the way home. To be rewarded, also on the spot, some “pips” for such efforts. Anything, the reader can think or have already thought in the past, can be rewarded with a small amount on any digital coin available today.

Why not try right now, and help this author deliver better and more often, content like this in your mailbox, while at the same time, you, the reader get an opportunity to explore and learn the advantages and benefits of doing so? Today the minimum amount being asked is 1 euro and one euro only! Click on the button below, the one it says “buy me a coffee“ and afterward don’t forget to leave me a tweet @AeonLabsS.

Until next time, don’t forget to share, and if possible, subscribe for fairness of work and a fairer and safer economy.

a #tag can be an important keyword to someone reading it, unlock an area of search results and knowledge

This particular article took 2 hours and 40 minutes to be written, top to bottom

This article is looking for sponsors to grow. Consider giving a donation by Paypal to the author using the email mtpsilva@gmail.com or at least buy him a coffee by clicking on the above image link.

As part of my research on online live writing, no proofreading was made to this text. Only Basic automated corrections. Over time the reader can expect additional changes to the text without prior notification: bugs and error correction, but also for content adding. So make sure to check back later the article.If the reader, instead, prefers reading well-formatted text paragraphs, error-free; perfect notations; high-resolution graphics, figures, photos, and videos, please consider subscribing on substack.com to a monthly or yearly plan. To do so, the reader can start by clicking on the “subscribe now” button above.

this won't take too long...

Miguel Tomás — Wed, 30 Jun 2021 14:45:52 +0000

my computer is like this for the past 2-3h or so....

"this won't take too long..."

Now Available to download on Google Play

Miguel Tomás — Tue, 29 Jun 2021 16:54:59 +0000

Now Available to download on Google Play AeonLabs app for remote management of stock materials and products

https://play.google.com/store/apps/details?id=aeonlabs.cloud.barcodescanner

The 30 and cents, Durable, Easy Repairable on Site Remote Monitoring of Concrete Maturity

Miguel Tomás — Wed, 23 Jun 2021 13:44:15 +0000

The other day, someone whispered to me, here in the neighbourhood where I live, “why not develop your own“ maturity sensor device, after I ended a short chat session on WhatsApp about it. At that time it was meant to be a simple chat conversation starter on WhatApps with my long-lasting research colleagues in science.

A week went by and I continue doing what I have initially planned: modeling in Tensorflow some research data I have here, from my 1st masters’ thesis about self-sensing carbon fiber composite sensors for IoT applications. During that time, I got some time to browse what is already on the market, for monitoring concrete maturity in real-time and with live data. I found the following enterprises already commercializing IoT devices with that specific purpose:

continue reading below
https://migueltomas.substack.com/p/the-40-and-cents-durable-easy-repairable

Modelling in science is like...modelling elsewhere

Miguel Tomás — Tue, 22 Jun 2021 14:55:05 +0000

1 week later of ML modelling...

Miguel Tomás — Tue, 22 Jun 2021 12:23:52 +0000

1 week later. 168h uninterrupted on intel I7 with 16Gb of Ram. here it is the result in just one excel chart

Scientific.net Number of reads WebHook

Miguel Tomás — Mon, 21 Jun 2021 12:53:24 +0000

In the meantime i've coded a small webhook to retrieve the number of reads from a paper published on scientific.net and insert it to any author/ researcher webpage:

HTML frontend code:

<script src="js/jquery-3.4.1.min.js"></script>
   <small id="include-aef-stats">
         <script>
             $(function(){ $("#include-aef-stats").load("kernel/loadstats.php"); });
         </script>   
    </small>

loadstats.php backend code:

<?php
$url = "https://www.scientific.net/Paper/GetDownloadsAndVisitorsCount?paperId=523460";
$scientificNetCounter = file_get_contents($url);

echo $scientificNetCounter." reads on A.E.F. website" 
?>