Carina by Hellmut1956 - scale 1:20 - long keel yacht around 1900

[Hellmut1956]

Due to the limitations of the number of pictures I will complete for now this thread of actions and continue later with the electronics. Here some pictures that might be of interest and show the function of the aluminium bars for the aluminium sheet.

 

 

In this picture you can see the outer and the inner wall of the battery box with the aluminium bars. You can imagine seeing that picture how an aluminium sheet will be surrounding the batteries and how the aluminium bars will support the aluminium sheet. Of course I have to use the fake batteries as the electrical isolation are not in place yet.

[Hellmut1956]

Finally I have the second inner wall, now with the aluminium bars in the final length. I decided to make them 2 mm longer to ensure that never both poles touch the outer walls. Just one more provision the prevent shorts!

 

 

Now on the front inner wall, the one at the bottom, I have to mill out the inner parts as done with the other one. Funny was the feeling to have something so solid that fits into my hull!

Edited February 20, 2013 by Hellmut1956

[Hellmut1956]

Hi friends.

Since the last time I have published something about my project a lot has happened. The biggest factor slowing down the progression of my project has been my health, followed by topics, fascinating for me, that have opened to my eyes new fields to pursue as part of my project!

2 strokes, 1 thrombosis, responsible for making my second stroke less strong, a couple of times my heart stopped beating due to side effects of my medicine for lowering my blood pressure. Due to the thrombosis that took place due to the cables inserted in my veins that connect my pacemaker with my heart, I had to take a special medicine to eliminate the thrombosis and this accidentally also helped to delude my cerebral infarct. Really the biggest effect on my daily life and capabilities to allow me to work and study for my sailboat project was one of the times my heart stopped beating. It caused damage in my brain. My studies on the other hand help to execute my muscle between my 2 ears helping is to reorganize itself to overcome the impact of the damages. Since my heart events I have never again be fully operational! This is not to have you feeling compassion for me, this is to share that the fascination of our hobby helps to combat the problems due to health issues to those that might share health problems with me!

 

As I started to report in my other thread about modelling my sheet control system, my focus moved from hardware oriented work to such of software and electronics, as well as skills development and improvements on science areas related. The so called MOOC courses, free online courses offered by the best universities in the world through sites like "Coursera", "edx" and so on I have been refreshing my knowledge in mathematics, physics, electronics and IT.

 

The starting point was my activities towards gaining the ability to model by sheet control system. When you model some system you need to verify the quality of the model by comparing the results of using the models with the real world. I decided to use RaspBerry Pi´s, as this electronic board is being supported by the software from "Wolfram Software", Mathematica and SystemModeler.

 

So I did start by making me familiar with the Raspis, to learn how to control them from my desktop PC. A RaspBerry Pi with a WLAN module, all very inexpensive, Raspi 3B even has Bluetooth and WLAN integrated, make it possible to access the Raspi via my local WLAN and even via the Internet. Using the proper means you can control your Rasp board from your PCs desktop, keyboard and mouse, as if those would be directly connected to the Raspi and have its desktop displayed in a window on your PC desktop! I got fascinated, that using a Raspi in a naval model you could communicate with it via wireless, be it WLAN, Bluetooth or cellular. Linked to this was that I did find a provider, that offered cellular phone service at 1 Euro per month and 1 additional Euro per month for 100 MBytes of data. You could even book more data volume for single days. This way my model sailboat in my workshop could be controlled for free using my in house WLAN and up to 100 MBytes per moth if on a place without WLAN availability!

 

I did even register my then 2 Raspis, each with its own entry on a DNS service that is available for free!

 

But this did kill my setup! I have my PC always up-to-date on os and security software. But via my Raspis outside attackers did render my PC with Windows 7 Ultimate inoperable. All my efforts to repair my PC did fail. So I decided to upgrade it to Windows 10 Pro. That made my PC 100% operable again! But this painful experience that did take about a month to fix did make me decide to learn how to operate my setup as secure as possible! Now the Raspi operates with an Linux os called Raspian. So I started to learn about Linux.

 

But today's technological revolution around the Internet of Thinks, IoT, has had as a side effect intensive developments in the area to protect so called "connected devices" from cyber attacks. Even the different states are starting to implement laws that enforce people developing IoT devices. Recently a US government agency has files a suit against a Taiwanese company for not securing the products against cyber attacks. Just keeping track of the explosion of activities to protect connected devices from cyber attacks is beyond my abilities to learn. So I have limited myself to those developments in the area of ARM based devices, as both my Raspis, as well as the other electronic sub systems use ARM Cortex based devices. A result of this has been my definition of an IT setup in my workshop:

 

 

Basically the picture shows a user like me, sitting in front of a screen and interacting with the experiment setups via WLAN/LAN/cellular.

 

When you go into the details of Linux and how the communication to local/remote hardware takes place, you find out how to make setups to reduce the risk of being attacked successfully. The potential for attack is expressed as "vulnerability area". A lot is achieved by changing the execution of functions from default to unusual setting and to limit who is authorized to communicate and to limit what an authorized entity is entitled to to the minimum required.

 

But those performing cyber attacks are getting more and more sophisticated. Recently the German defense minister said that one of the mayor areas Germany is extending their spending to a "cyber division" of the German military. So the area of cyber war is an area were heavy investments are made by governments. But also all of the other fields of technology that are required to participate on the global digitalization, not only IoT, but also Industrial IoT, called IIoT , medical, automotive, infrastructure and home automation  have to jump onto the Internet. So very deep pockets are involved in developing protection to reduce vulnerability, in developing cyber war methods. Also the research activities in areas related to this cyber war are developing the scientific ground for safer solutions!

 

It is interesting to mention, that no device connected to the Internet can be made 100% secure. all is about making the cost to succeed as high as possible by making the investment of extensive resources necessary! At least until quantum computers become reality, what as a result forces sciences related to cryptology work on finding methods of encrypting that are able to withstand attacks using quantum computers.

 

There is the technique of "virtualization" that is protecting environments in the computer server world. It is said that the methodology of virtualization is one that achieves 100% isolation and as a consequence protecting the environment which uses virtualization from what ever comes from either those virtualized environments of from attacks to such virtual environments. Intel CPU's have support for virtualization build into their hardware and newer CPU's do have more sophisticated support for virtualization. So in the setup for my workshop shown in the graphic above the computer is setup so to benefit from virtualization! This methodology is available since a long time and can be seen as mature. The challenge is to make it possible to use this methodology in deeply embedded systems. We as naval model aficionados just call it "electronics" and mean simple boards using a micro controller, ARDUINO is well known as is RaspBerry Pi.

 

 

This small graphic presents the 2 types of so called "Hypervisor" that make the virtualization available in Systems. The "Hypervisor Type 2" is an application that runs on an os like Linux, Windows Server or Desktop. Here the hardware of the system is controlled by the os as we know it from our desktop PC's and on it is running an application called "Hypervisor", "VirtualBox" being an example of the many products available. The hypervisor type 2 the opens guest applications with their own operating system. The os in this environments believe they are the only software running on the hardware so that unmodified operating systems and application can be executed. No event taking place within this environments can access the hardware directly. The hypervisor intercepts them and returns to them as if they were doing it on their own. This makes it possible to run i.e. Linux Ubuntu on a PC with Windows Desktop. The draw back of this is that it requires extensive resources, as it has to duplicate a complete OS for each virtual environment! Also an important draw back is, that it is impossible to achieve predictable real time responses. Known should be this setup for anyone who rents a server from a provider. He does not get his own physical server but just a virtual copy of it!

This virtualization ensures 100% isolation between the virtual environments and its applications.

 

On a board with a controller on it we have to basic versions. Arduino is one, where the programs we write ´for it are executed directly of the hardware, it is called "bare metal" . Arduino have very little flash memory and even far less RAM. So here a hipervisor type 2 is basically impossible to use!

The issue is different for boards of the type of RaspBerry Pi's. They have still limited hardware resources compared with Desktop PC's and Servers, but enough to allow a Linux and its application to be executed on them. So hypervisor Type 2 are possible for them and as long as our application can live with those limited resources, all is OK. But still real time is only possible to implement in a very relaxed way!

 

So lets look into Hypervisor Type 1. This software is executed as we introduced the term earlier here, running a a bare metal software. On a hardware of the kind like the RaspBerry's  hypervisors Type 1 are meaningful and due to the fact that the vulnerability issue is getting a mayor aspect you can find ports of Hypervisors type 1 of the Raspi's. What remains affecting such systems is their hunger for resources, but using a hypervisor type 1 or 2 on a Raspi is possible.

 

 This picture shows what is demanded for any comercial application that has access to the Internet. The topic of security and vulnerability start from the very beginning in the Design phase and goes to its desposal or destruction. Like it is happening with the drones for our aviation model makers, law will enforce more and more rules to be able to fly drones and at one time this might extent to naval modeling R/C. I am a believer that providers have to develop setups that hide the complexities of reducing vulnerability, for commercial developers and even more for hobbyists!

 

A first result of my reseach work to the topics listed so far is my need to develop a workflow which uses a development setup on my PC that makes the development work to take place in a virtual environment. Further it requires me to understand how to develop from such a virtual environment for my RaspBerry Pi boards, for my Arduinos, for my other ARM based boards. This requires to fully understand and execute my development work in this way. There are ways to do it, but I need to start experimenting with it. Here my health problems are blocking me to develop a clear picture of how to do it and to start doing it. I will need to setup a checklist that achieves a safe and effective setup. I do my software development using the programming language "Python" and the development environment, short IDE, PyCharm from Jetbrains. 

 

But, typical to the speed technologies are evolvong, there is a second technology besides "Virtualization" called "Containerization"! Here the graph comparing both technologies:

 

 

 The graphic of this compare of the 2 technologies make evident why "containers" are better suited for embedded systems! At he initial phase boards of the kind as RaspBerry Pis are more adequate that those like Arduinos. The dominant but not only tool for this technology is called "Docker" and is available for free, You can also see that Docker takes the place a "Hypervisor Type 2" takes in a virtualization setup. But you can also see that Docker based applications do not require their own copy of an operating system, but just those services that are specifically required to execute a certain application. While virtualization offers 100% isolation and so reduces the vulnerability of virtual environments and the host os, the isolation in a container environment is more equivalent to how multiple applications that run on a desktop OS like Windows. We know that if one application crashes there is a fairly high probability that the whole system crahes or at least gets damaged! So the next step of the research taking place this days is to develop a hybrid solution that combines the advantages of both methods while reducing as much as possible their disadvantages. To better understand the line along which such a "hybrid solution" is probably to be found it is helpful to know the concept of "micro services"! Micro services are a method to implement software applications by dividing it into sub functions are the so called micro services"! An application so becomes a bunch of micro services interacting to deliver the functionality of the application. The benefit of such a way to program applications results from the encapsulation of software functions. Each such software function is implemented in such a way that it interacts with other micro services via a common means of doing so, so that errors cannot propagate. This makes it dramatically simpler to ensure a certain functionality is bug free and implemented in such a way it is reusable to the maximum extent possible. Corporates can that have those micro services be distributed in their corporate environment in a way to ensure best load over the resources of a corporate.

 

So if we envision containers to implement such a "micro services" kind of functionality this would benefit that it could be executes were ever a container environment is available because each container includes all it requires to execute the code within it. So containers are used "orchestrating" them to implement a functionality desired. Each of the containers allows to define what other containers are allowed to interact, how they interact and to what each of this interaction is entitled to. 

 

So while I am still working on getting a more complete understanding of Docker and its set of tools to be able to add from those understandings were and how I can define who is entitled to what! My objective as of today is to do all development work from a virtualized environment to reduce as much as possible the so called areas of vulnerability on my PC and on my RaspBerry Pi. Only when i have such a check list I will start to do my experiments required to verify models of my sheet control system for my sailboat model "Carina" and to use the same scheme to have the electronics in my sailboat communicate as well as my sailboat to communicate with my radio control for the sail boat!

[druxey]

Helmut: I 'm sorry that you've had major health challenges, but you seem to have overcome many of them! Hopefully things are now stable and improving, and that you can continue your modelmaking rather than health concerns.

[Jim Lad]

It's good to hear from you again, Helmut.  It seems that you've had some really major health problems in the past and it's great to hear that you've been able to 'get back on your feet' again.

 

John

[Hellmut1956]

Thx! I will probably never get well again. I had to decide I will never again be able to fly to Bogotá, Colombia, where I did live before relocating to Germany in 1976. But I do not complain and i am happy that my curiosity and fascination for our hobby in general and the way I am working on my project is more pushing me to overcome limitations that my health is imposing! I kind of do get angry with myself when I am not able to concentrate enough to continue one of the many streams of work I am pursuing. Also I am more then just lucky to have a wonderful family, a loving and supporting wife, 2 girls, 25 and 27 years of age and my boy with 23.

 

But my activities to take benefit from the technological revolution taking place this days for my model are my biggest motivators. I am still working to improve and organize my workshop. I do plan to share my work on it with you.

[janet bode]

janet B

 

found your article very interesting Hellmut love the shape of your hull.I too became very unwell but i

 

find that watching other builds of all the members a great lift up for me,anyway best of luck with your project.

[Hellmut1956]

Dear Janet, it is key to help our body to overcome or at least reduce its problems when an activity lifts our mood! 

 

I just had my 4th stroke. Fortunately it was now for the third time in a row a light one called TIA. My big problem right now is that am most of the day very tired and so much sub projects I want to get ahead with. Also on my list is to update this report.

 

But most of my activities are related to advance my workshop, where electronics play a big role!

Edited September 12, 2018 by Hellmut1956

[Hellmut1956]

Just coming back from a stay at the hospital for a week due to diabetes problems. This stay was useful to read, was useful to reflect and to think about the possibilities to apply what I learned in those readings to my model. 2 "topics" did come out of this exercise. One was the possibility to create an Alexa speech interface with the electronics in my model. The other one was to investigate if artificial intelligence could be applied to my sheet control system.

 

Both of this activities do require to think about what, how and if meaningful it is to apply this.

 

So for the Alexa interface, voice interface does require to think about where such an interface would look efficient or at least funny. So i.e. setting up the model at the shore of the pond or during the tests in my workshop. This does require quite a bit of work to analyze it.

 

The use of artificial intelligence as part of my sheet control system requires me to reflect what "inputs" would be valid and against which the AI would do the fine tuning of the sheet control system. Another 2 true inputs are the position of the control stick on the R/C transmitter that defines what is the maximum angle of opening the boom is allowed to have. The other valid input is the one coming from the magnetic angular sensor keeping track of the boom. This is the true actual angular position of the sail.

 

The use of AI requires to have a lot of information, called datasets or examples, to train the AI algorithms. As my sheet control system uses a couple of electronic sensors, each of them can generate thousands or hundred of thousand data points as the sensors do provide data to the system at pretty high frequencies. It is relatively simple to generate those examples in my workshop, as a ventilator would imitate the wind and apply pressure to the sail. The challenges related to this concept are on one side related to identify the different possible or likely scenarios and as a consequence the type of actions a KI would have to do.

 

The other side is to think which method of AI offers the functionalities I want. One aspect is that the AI should continue to learn when rolled out. So I am planning to define and specify what I want the AI to do and what data points are available so to identify which methods of AI could be appropriate. With this, I want to ask in AI related forums for support.

[mtaylor]

Hellmut,

I hope the hospital time helped your health.

 

Good luck with the project.  AI is still pretty new and experimental.   If you can get the memory and processing down to small enough physical size, you might be onto something.

[Hellmut1956]

Finally my values of sugar in my blood is getting fine. Sadly the health problems have revived a bit the problems I had after my heart did stop a couple of times. After a while my brain tells me it cannot continue. I also have problems self motivating myself, something that wakes up my will to fight against those limitations.

 

What I usually do in the morning is getting into my workshop, go through mail and then play some games on my computer. By now i am bored about playing those games and so I am forcing myself to continue work to get order into my workshop. What I am proud about are the improvements at my workshop office space:

 

 

My graphics board in my PC, GTX1060 with 6 G Byte of memory, is no using 4 monitors connected to it. The lower 2 are 28" UHD monitors, the upper one on the right is a Samsung 24" monitor, the one to its left is a 23" monitor, Additionally I have purchased and connected the Logitech Z607 5.1 Sound Boxes. Was not trivial but finally I got it done!

[Hellmut1956]

@Big Rat: I need to revive my enthusiasm to research, experiment and to work on improving my workshop. Regarding AI there is a development to implement what is called "Neuromorphic System"  The company BrainChip has developed the functionality into a device called "Akida". This works in a way called pulsed and makes it possible to implement in devices connected to the Internet, called etch devices, to run AI acceleration which is orders of magnitude more performant than existing implementations and demands less power, again in the order of magnitude. I guess phones that have this kind of accelerators for AI will begin to be available in 2020. So I am in no rush!

[Hellmut1956]

Another project pending to be completed is the next generation panel for my workshop:

 

 

This is what I do call may Panel 2nd Generation. Above this panel you see the little displays I found at a Chinese source that do display voltage and current values. I thought it would be great to have one of this displays for each voltage my panel is supplying. What I learned later was that to have those displays show the current voltage and current for each voltage supplied by the panel I had to implement mayor changes.

 

 

I do not enter in details as this goes pretty deep into electronics. What you can see is that now each supply connectors now consist of a pair of connectors, both poles. You can also see the switch available at each voltage source. I am replacing them by ones that do emanate red color light when in the "ON-Position". To the right is the opening for a 5 mm diameter RGB LED which will be configurable remotely. At the top you see now the small displays were they belong to and at the bottom screw able connectors. Both them and the connectors above the switch are switched ON/OFF by it, the 2 pairs below the switch are constant ON ones as soon as I switch my modified PC power supply to feed the panel.

 

 

As it happens usually one change leads to more changes. I have to develop and make small PCBs to have the circuitry to feed the RGB LED.:

 

 

Each of the 3 pins to control the color that will result displayed from the LED next to a switch. You can appreciate how small those surface mounted resistors are.

 

Each one of the 3 pins of a RGB LED are driven by something called PWM, pulse width modulation. This is used to control the amount of current flowing thru the LED. 8 RGB LED, each with 3 pins that require a PWM to control the amount of current results in 24 PWMs.

 

This board contains an IC, the PCA9685 that is capable to generate 16 PWMs. So 2 of this boards, total 32 PWM are required to generate the 24 PWM I do need.

 

 

This small RaspBerry Pi board, its latest revision offers WiFi and Bluetooth connectivity. It will program the 24 PWM so that each RGB LED displays the color code related to the voltage being supplied. I will be able to define and change the color of each LED from a window on my PC that shows the Linux desktop running on the RaspBerry Pi.

 

As you can imagine, the delay of my ability to work also on this project due to my health problems makes me angry with myself and it has to be the resource to overcome the limitations due to my health problems.

[Hellmut1956]

Hi my friends.

It has some benefits that I am not progressing with building my sailboat due to health problems. reflecting things about the design of the sheet control system presented here made me aware of an aspect that could be addressed thanks to the intelligence of my sheet control system.

 

 

My sailboat model shown in the photo has beams on main and fock sail. I do plan to be able to fully open the sail, 90° to the hulls main axis, to fully benefit the wind coming from the rear. But this has a problem! The beams would touch the water surface when the hull gets away from the perpendicular position. So opening the sails up to 90°, as it is my objective, I do need to ensure the beams do not touch the water! So there is a maximum inclination degree were the beams touch the water surface.

 

So my sheet control system has to take this into account by having a sensor that keeps track of the inclination. The input from this sensor has to change the maximum allowed opening, which usually is managed by the position of the associated stick on the remote control, to ensure the beam s do not touch water!

 

As you can well imagine, the stepper motor used as winch has no problem with this duty, the computation in the microcontroller to monitor the inclination sensor, is also fast enough. But what is with the mechanical part of it? How fast is it possible to pull in the sheet to pull in the beams? Here experimenting will be key. But also intelligence might be able to include what is inclinating the hull and ensure a maximum pull speed is being taken care of. The cause can be either the operator deciding to change the direction the boat boat is navigating or a change in the direction of the relative wind angle is. In both cases the intelligence and the mechanics of my sheet control system can and has to ensure the inclination of the sailboat does not result in pulling in the sheet faster than the maximum speed to be defined.

 

Fortunately acceleration and angle sensors are cheap and extremely powerful thanks to their use in smartphones.  

 

I know, my sailboat model will, with a very high probability, never finish and sail before I die.

[Hellmut1956]

The last years I have focused on improving my workshop and that still keeps me busy. Let us see hov the future will go.

[Hellmut1956]

 

I am still alive and the focus and focus my activities on my workshop.  Will write a report soon. This image shows The monitor setup that I have on my "office" desk in my workshop. The reason for this is the help of an old friend with whom I use to fly his real airplane, a Bonanza 35B, in California, Nevada, and Arizona. He did help me to get a brand new PC with an RTX 3080 graphics card and flight simulation peripherals from Logitech. So while I profit from this setup when I return to the electronics in my sailboat model it is also my virtual "cockpit" for flights in the Microsoft Flight Simulator 2020.

 

Also, my workshop has a Creality Ender 5 Plus 3D printer for which I build a platform on wheels similar to what I did build for my circular table saw and milling table based on a Bosch GTS XC Professional and a Triton TRA001:

 

For the circular table saw I build a system to have the altitude of the circular saw disk above the surface of the table. This way I can set the depth the saw cuta into a wooden plate by 1/100 mm presision. I am planing to make the same for the milling table for which you see the opening where the Triton TRA001 is placed.

 

 

 

As the Triton TRA001 router is pretty heavy I have to build support that takes the weight of the router. The other related project in the aspiration system for my workshop and to which I have connected the table saw. As soon as I have my 3D printer in shape to be used I will have to print 3D a device that allows the aspiration system to also get the dust that the table saw generates above the saw table surface.

 

 

 

 

 

 

Edited March 24, 2023 by Hellmut1956