Milan Vukov

📣New Course Offering: Simulink Real-Time HDL Workflow with Speedgoat Hardware The two-day course focuses on partitioning Simulink models intended for real-time execution on Speedgoat target machines to execute on the CPU and FPGA for Rapid Control prototyping (RCP) and Hardware-in-the-Loop Testing (HIL). 🔗click to register https://spr.ly/60415rLb3

2

Interactively create AI-based reduced order models (ROMs) of subsystems modeled in Simulink using the Reduced Order Modeler app. Using this app, you can create ROMs from full-order high-fidelity third-party simulation models brought into Simulink as Functional Mockup Units. #reducedordermodeling #Simulink #AI

1

Next Tuesday 8th October 2024, I will be talking on the Belgian C++ User Group gathering (https://lnkd.in/evHVJB_x) Title: Better Threading in C++20 Abstract: We got threading support by the language and library in C++11. There were however a few things missing, like RAII for a thread object, or a nice way to stop a thread, or to request it to stop. Similar problem, how do we stop a ‘waiting’ condition variable. Most developers have solved these problems with home made solutions. C++20 brings standardized solutions to the table. Next to that we also get some new synchronization facilities like barrier/latch/semaphore, all nice additions. There is no longer a need to re-invent the wheel or reach out to 3rd-party libraries. We will look at all these new little tools and investigate on how to use them. So register your stop callback, and by the end of the talk we will request our source to stop, and in our ‘cooperative cancellation’ scheme, do not forget to keep an eye on your stop token. I am looking forward to meet you there. In case you are looking for C++ training in your company, feel free to contact me.

21

📣 Students: Registration for the Simulink Student Challenge 2024 is now open! Win up to $1,000 (USD) by submitting a short video that shows what you’ve accomplished using Simulink. Learn more: 🔗 https://spr.ly/6047qR5YS

1

Learn the basics of simulating a simple battery management system (BMS) for safe charging/discharging in various temperatures and more in Simscape Battery Onramp. #electrification

1

👉Attn Engineers: Join us on May 29th to learn Simulink basics for modeling, simulating, and testing control systems. This webinar also covers how to tie requirements to your design, test for traceability and verification, and generate code for your software implementation. Register today! 🗓 #simulation #modelbaseddesign

1

💡𝐇𝐚𝐫𝐝𝐰𝐚𝐫𝐞 𝐢𝐧 𝐭𝐡𝐞 𝐋𝐨𝐨𝐩 𝐒𝐢𝐦𝐮𝐥𝐚𝐭𝐢𝐨𝐧 𝐚𝐧𝐝 𝐓𝐞𝐬𝐭𝐢𝐧𝐠 𝐂𝐫𝐞𝐝𝐢𝐭: www.halvorsen.blog 🔧 𝐄𝐱𝐜𝐢𝐭𝐢𝐧𝐠 𝐏𝐫𝐨𝐣𝐞𝐜𝐭 𝐎𝐯𝐞𝐫𝐯𝐢𝐞𝐰: As System Engineers, we love a good challenge! Recently, I had the opportunity to delve into an intriguing project outlined by Hans-Petter Halvorsen on his blog ℎ𝑎𝑙𝑣𝑜𝑟𝑠𝑒𝑛.𝑏𝑙𝑜𝑔. The project involved developing a Control System Prototype/PoC using Hardware in the Loop Simulation, a crucial aspect of the development and testing process. 🛠️ 𝐄𝐬𝐬𝐞𝐧𝐭𝐢𝐚𝐥 𝐂𝐨𝐦𝐩𝐨𝐧𝐞𝐧𝐭𝐬: In this Proof of Concept (PoC), he utilized a Fuji PXG5 PID to control an industrial small-scale process. The HIL System was implemented using LabVIEW, showcasing the versatility and power of this platform. 👩‍💻 𝐔𝐬𝐞𝐫-𝐅𝐫𝐢𝐞𝐧𝐝𝐥𝐲 𝐈𝐧𝐭𝐞𝐫𝐟𝐚𝐜𝐞: A key aspect of the project was creating a user-friendly GUI/HMI, ensuring ease of use and understanding for operators and engineers alike. 📈 𝐑𝐞𝐬𝐮𝐥𝐭𝐬 𝐚𝐧𝐝 𝐈𝐦𝐩𝐚𝐜𝐭: The delivery included a comprehensive Technical Report, detailing the system, methods used, and the results achieved. This project is poised to be a cornerstone for future decision-making within the company, demonstrating the value and innovation we bring as System Engineers. 👍𝑯𝒊𝒕 𝑳𝒊𝒌𝒆, if you found it helpful !✨ 🔁𝑹𝒆𝒑𝒐𝒔𝒕, it to your network !✨ 🔖𝑺𝒂𝒗𝒆 it for the future !✨ 📤𝑺𝒉𝒂𝒓𝒆 it with your connections !✨ 💬𝘾𝙤𝙢𝙢𝙚𝙣𝙩 your thoughts!✨ #HILsimulation #Labview #FujiPXG5 #PID #ProofOfConcept #simulationtechnology #engineeringinnovation #Labviewprogramming #Fujiinstrumentation #PIDcontroller #simulationtesting #engineeringdesign #Labviewdevelopment #FujiPXG5PID #ProofOfConcepttesting #simulationsoftware #engineeringtechnology #Labviewautomation #Fujiinstrument #PIDcontrolsystem #HardwareInTheLoop #TestingAndValidation #PIDControl #ProofOfConcept #SystemsEngineering #RealTimeSimulation #QualityAssurance #EngineeringExcellence #ProductDevelopment #LabviewProgramming #ControlSystems

206

2 Comments

A welder can't carve circuits into a silicon chip with tools like this. Precision tools from companies like ASML can carve chips, because they work at a nanometer scale. The right tool for the job. We see the same thing in inter-company networking. Every S2S VPN, B2B VPN or extranet is a welder trying to use a blowtorch to do silicon chip lithography! VPNs and extranets are the wrong tools for the job. This is why enterprises are now telling partners to take their blowtorches and go home. The massive cybersecurity risk of providing network access is becoming a dealbreaker. Fortunately, there is a new option. The right tool for the job. Lasers instead of blowtorches. NetFoundry's label for these lasers is AppNets. AppNets securely connect individual workloads. Your networks no longer need to connect to other networks. AppNets already connect billions of sessions per year, in places like the world's largest banks and critical infrastructure. Ping me if you want to be next to trade a blowtorch for a laser.

38

4 Comments

Our latest paper on Linux in Automotive, written for the „ATZ Automated Driving“ conference is now available at Springer.

9

Hi folks, as a seasoned embedded systems developer, I've seen the pendulum swing back and forth between using Real-Time Operating Systems (RTOS) and programming in bare metal. While some still swear by the latter for its supposed 'lightweight' efficiency, I firmly believe that RTOS is the better choice - every time. Here's why: - Simplified Development: With a RTOS, you can focus on your application code without worrying about low-level details like memory management, interrupt handling, and synchronization primitives. This leads to faster development times, reduced bugs, and improved overall quality. - Better Multitasking: RTOS provides robust multitasking capabilities, allowing multiple tasks to run concurrently while still meeting real-time requirements. No more tedious task switching or context switching - your code can run smoothly and efficiently. - Increased Reliability: By leveraging a proven and well-maintained RTOS kernel, you benefit from the collective expertise of thousands of developers who've contributed to its stability and security over the years. - Future-Proofing: As software complexity increases, a RTOS provides a solid foundation for scaling your application without sacrificing performance. No need to rewrite your code or refactor your architecture - simply add new features and tasks as needed. Don't get me wrong; bare metal programming has its place in specific, niche applications where overhead is a concern. But for most use cases, the advantages of using a RTOS far outweigh any perceived benefits of going bare metal. So, what's your experience with RTOS vs. bare metal programming? Share your thoughts. #RTOS #EmbeddedSystems #SoftwareDevelopment 

83

7 Comments

I've been discussing about this recently with the team at Thinking in code: Is using floating point arithmetic a good idea in projects with high ASIL levels and strict security requirements? Floating point math is common in software development, but when we're dealing with safety-critical and security-critical automotive systems, it's not that simple. Predictability Is Crucial In safety-critical systems, we need things to be predictable. Floating point operations can introduce rounding errors and unpredictable behavior because different hardware handles these calculations differently. We can't afford unpredictability when safety is on the line. Verification Challenges Proving that floating point calculations meet safety requirements is tough. There are so many possible inputs and states that testing everything is nearly impossible, which makes verification and validation harder. Resource Constraints Floating point operations can use a lot of processing power, which isn't ideal for embedded systems with limited resources. This can impact system performance and responsiveness—both critical in safety applications. Security Risks Errors in precision and rounding can be exploited by attackers, creating potential vulnerabilities. In high-security projects, we need to be extra careful. So... What Are the Alternatives? Fixed-Point Arithmetic It offers predictable behavior. In my experience, it's often preferred in safety-critical applications because it's easier to verify and validate. Scaled Integer Math Sometimes, we can get the same results using scaled integers, reducing complexity and increasing reliability. In projects where safety and security are so important, minimizing uncertainties is key. While floating point arithmetic might be necessary in some cases, we should carefully weigh the benefits against the potential risks. If we decide to use it, we need to: Test Thoroughly Implement rigorous testing to cover as many scenarios as possible. Use Certified Tools Use compilers and libraries that are certified for safety-critical applications. Document Everything Clearly explain why we're using floating point arithmetic and how we're reducing the risks. Have you faced this issue in your projects? Did you choose to use floating point arithmetic, or did you find an alternative? I'd love to hear your experiences and thoughts!

15

3 Comments

A long-delayed post about importing the standard library in C++ using the latest CMake 3.30 release and a recent Clang 18.1.8 toolchain. It works (mostly) and continues to improve! I linked to a few accompanying commits where you can see the details too but I only tested it on an up-to-date Arch Linux install :-)

22

I am really looking forward to taking part in this webinar, I love FPGA development and taking a model based approach is something which enables us to deliver on time, budge and cost. I will be sharing several real word examples of where a model based approach has been successful. #fpga #embeddedsystems #electronics #embeddedsystems #engineering

189

3 Comments

Hi folks! Microkernels are gaining traction in embedded systems, but are they worth the additional communication overhead? Let’s break it down: Pros of Microkernels in Embedded Systems: ✅ Enhanced reliability, as faults in user-space services don’t crash the kernel 🛡️ ✅ Improved security by isolating critical services from the kernel 🔒 ✅ Greater flexibility for adding or updating system components 🚀 Cons of Microkernels in Embedded Systems: ❌ Increased inter-process communication (IPC) overhead, reducing performance 🐌 ❌ More complex initial design compared to monolithic systems 📏 ❌ Requires more memory, which may be scarce in resource-constrained devices 🔋 When to Use Microkernels in Embedded Systems: 🔹 Perfect for safety-critical applications like medical devices or automotive systems. 🔹 Less ideal for ultra-resource-constrained IoT devices requiring high performance. What’s your take on microkernels for embedded designs? Let’s chat below!👇 #Microkernels #EmbeddedSystems #IoT #RealTimeOperatingSystems #TechInnovation

20

4 Comments