 Dobrej, sreči, da sem tukaj. So, sem Matas Jalardi z Yogitec. Sreči smo vse in smo partneri v ST. Yogitec je spesializati na funkcionalne zapečenje, vzvečati na vzvečenje in vzvečenje. Vzvečenje vzvečenje, kako je STM32, ki ne značila na funkcionalne zapečenje, prejzivno, da je izgleda po vsej zamočnih barbe in najboljih zelo potrebenih projekte. Poveno, da je zamočno zelo potrebenih barbe in tudi da je prav načinj, da se je zelo povedo, da je zelo potrebenih barbe in da je zelo povedo in pečne. Tudi, da je zelo potrebenih barbe in težko, in prišlično je vseh vseh izgleda. Vseh je izgleda IC61508. To je vseh vseh vseh. Zato je vseh vseh izgleda, vseh vseh domenja. Vseh je vseh vseh izgleda, vseh vseh vseh vseh, vseh vseh vseh, vseh vseh vseh, sboati vseh izgleda. Tukaj sem bilo našli, zvukaj sem zvukaj zjeba. Neljeste ste vseh zvukaj, zato vseh vseh zgleda. Zvukaj sem vseh vseh vseh vseh vseh vseh zgleda. Vseh vseh, vseh, vseh vseh ineliziru, vseh, vseh, vseh, vseh, That must be fulfilled in order to be used in a s2 or s3 application. Some of the requirements are already covered by the hardware, even though he was not designed for functional safety. Others are really application specific but they have to be covered by the end-user. z editečnim všgečjanjo. Treba je, da je vseglas očetke, začneva slajba, ili na srečih naši aplikacij, bil gledan i nekaj način v modneje. Tako, a issues na načinu, ta se je jemljate nari dela in začneva. Vsem lahko predelt prošeljem in tudi pošeljem in z editečnikem očetko, ki bomo zredne postočiti začnev, evo je nisi zrednji. Je tudi zgledal, da bismo opojo početke, to bil ki božna dela, da ne ne se prošelje. Če so pripravili, da se pripravili, da je kotek in kor, da bomo začinil vši data, da bomo dovoljali neko gačne, začinimo, da zelo se vse bolje obvediti, da so tudi vse vzelo vse, in da bomo vseo vzelo, da je je začinil vsi, da je to, da je tko, ki smo vzeli, še v Blackbox approach. Zato, ki so več zelo, je vsev, da je vsev, ki so tudi počal, ki so dovoljali, da smo vzeli vsev, da je to, ki so vzeli, vsev, ki so vzeli, of the STM32. If someone uses familiar on integrated circuits, so basically we receive the netlist and the RTL of the design. And based on that, we applied our analysis. So there is no guessing. It's really looking into the circuits down at the gate level. And then also we don't have to assume that the faults are equally distributed. In fact, we designed our library to focus on those parts of the circuits that they give the highest contribution in terms of failures. And, of course, this allows us to have an optimized software. And so we've done this using our methodology, which is a patented approach for analyzing integrated circuits. Really in a nutshell, this approach means that we take the netlist, we divide it into elementary parts. For each part we compute all the metrics. And then there is also a cell step which is about verification of the metrics using fault injection. So we inject all the faults at the gate level and see if the metric was accurate. So, yeah, so this is exactly what I was saying now. So basically using our methodology, we have this very accurate analysis in this case of the STM32. And basically from this analysis we derive the specification of our software and also the verification plan. What is this verification plan? Well, basically it's like a simulation. So we load the netlist and we have a golden image, which is the STM32 netlist without fault injected. And then in parallel we simulate the netlist with fault injected. So all faults are simulated. We run our software on top of that and basically we count how many faults we can detect. So, of course, this is a... Many of our customers, they try to do something like this by themselves, but then for them it's very hard to measure what is the actual diagnostic coverage, because you don't know, basically. In this case, basically we measure it and in the documentation, it's reported. So, of course, this means that you have a state-of-the-art solution with very accurate, you know, exactly what is the coverage contribution given by this library. So, about the partage with ST. So we have done this safety analysis and the first output of this safety analysis has been the safety manual of the STM32. This safety manual is now property of ST, so you can look at it, please contact ST people. And basically in this safety manual you will find all the list of the safety requirements that you have to fulfill, two or three. Then, of course, then at this point, on top of that, you have two options. You can either do your own development, you can do your own library, or you can come to us and get our library. And basically our library is testing the cortex and core, in flash, and part of the peripherals. So, basically, the library is doing that. You take the library, you plug in into your application, and after this integration, which is quite straightforward, that you can reach ST2, ST3 on your project. So, it's important to mention that these libraries are certified by Toofr Island, that they are absolutely application-independent, so usually we have almost no visibility in any kind of project, from robotic arms to heavy machinery to medical equipment, so it can be used on any kind of application. It's optimized in size and runtime, so the library is, well, just to give you a number for DF0 library, it's about 10 kilobytes of binary cores, so it's quite small. As I said, it's verified by fault injection, quantitative evidences of the coverage provided by this library. Of course, the design is modular, so if you want, you can run this library all at once and test everything in one go, but maybe you don't want to do that, maybe you want to you need to interrupt, you want to split the test in small chunks, this is possible. So, especially for the run test, which is the one that needs the interrupt to be disabled, so basically the test is organized into slices, so you can run all this, you can test all the run all at once or you can divide it into slices, so it's up to you how much of the run you want to test in one go. So this is to allow flexibility to the library. And then, of course, this library has been designed according to the 6508, but it's compatible also with other standards, so if you go through the documentation you will see how to map this library also in other standards. So, what is this library actually? So, we deliver the source code, we can see in the assembly, and we deliver the compiled scripts so far we support Kyle and the ARV, which are the most common used, sorry. And then there is the documentation of the library, so there is the user guide and there is the safety manual of the library. So basically you have two safety manuals, one of the microcontroller, which is owned by ST and then there is the safety manual of our library, and that we usually for evaluation we send to interested customers. So this is what we support right now, so we support the, we have libraries for the F0, F1, F2, F3, F4 and then one. I have to mention that we support all part number, so as you know STM32 has a lot of part number but we support all of them. And by the way we have a corner here inside the ST booth we have two demos, one running on the F0, the other one on the F3 so if some of you is interested please go and talk to my colleague he will show you the demo and also we can have a look at the code to understand how it works. And that's it. So basically, thanks to the software library you can use STM32 also for safety projects.