TABLE OF CONTENTS:
Turning the calendar over is an excellent time to review this year’s developments in the embedded sector and get an idea of what the future holds for the industry in 2023. Trends for embedded systems for 2022 might have already caught your attention. This time around we have a few embedded programming trends to watch in 2023.
Due to increasing global customer awareness and cybersecurity concerns, the following year (and beyond!) will pose enormous challenges for manufacturers and developers of medical systems distributed in the United States. During the 117th Congress of the Food and Drug Administration (FDA), H.R. 7667 legislation introduced requirements for cybersecurity in medical devices. A large number of requirements are imposed on manufacturers by this act, including a mandatory minimum level of I.T. security, provision of patches throughout the device’s lifecycle, or a summary of the software components (S.W.) used. However, cybersecurity has not only taken hold in the United States. The European Union is working on a new requirement regarding criteria for evaluating IoT devices for cybersecurity and putting such information in the form of pictograms on the packaging. Embedded, IoT and other areas are intrinsically linked to cybersecurity, regardless of whether we want it or not. As a result, the issue is becoming increasingly relevant in many industries. Manufacturers and distributors will need to adapt to the updated guidelines and requirements, which will require considerable funding and effort.
Despite the enormous problems with purchasing many semiconductor components, as well as the U.S. sanctions on Chinese semiconductor manufacturers, more and more products are appearing on the market. The shortage of microcontrollers is causing 8-bit platforms to return to favor with some manufacturers, as their availability is more widespread than 32-bit platforms. Meanwhile, others are using ESP32s from Chinese companies.
Although everyone seems to be doing their best to cope, the trend remains the same for the time being. There will continue to be component shortages in the market unless the global recession grows, reducing demand for devices.
In the hardware area, diversification is the trend. Some time from now, a certain amount of semiconductor production will be moved out of Taiwan to other locations such as the U.S. or Central and Eastern Europe.
The impact of the chip shortage on software production is an eye-catching phenomenon observable on the market. This manifests itself in the increased popularity of platforms such as Zephyr. This framework attracts an ever-increasing number of contributing developers yearly, with more than 300 people joining the project in 2022 alone. Some companies have already started porting their legacy solutions to this platform. The framework reduces the dependency of software on the hardware architecture (HW) of the processor, thereby enabling the firmware portability that is so sought after these days, which in turn results in a much faster process of transferring solutions to a new processor. We are talking weeks, not months, spent on code porting work using Zephyr. Sounds interesting, right?
Yes, and it is not a fledgling endeavor! The framework I want to tell you about makes it possible to create C# applications for embedded platforms. It’s interesting to note that their performance doesn’t need to be cosmic to run smoothly. So what is the name of this framework? This is the “.NET nanoframework”.
It’s worth noting that the solution is doing pretty well, as evidenced by the number of supporters and contributors (not as large as for Micropython, but still significant), demonstrated by the traffic on the repository and Discord. It appears that some companies use the nanoFramework in commercial solutions and praise this approach to software development. There is no doubt that C# has an extensive toolchain and a number of tools supporting software development, which gives it an edge over other programming languages. In addition, the framework allows inserts to be created from C++ code, enabling more deterministic handling of selected parts of the software. The list of platforms includes microcontrollers from ST, NXP, or Texas Instrument, among others. As a bonus, there are a large number of libraries for peripheral devices such as temperature sensors, humidity sensors, external ADCs, and lidars.
In an age of finding ways to quickly port solutions between platforms and a shortage of C/C++ developers, is this approach likely to catch on? We shall see. I’ll certainly be keeping an eye on this technology on my watchlist moving forward.
Among programming languages, Rust is gaining ground, notably in the embedded industry. The language is receiving special attention at developer conferences. Rust in Linux is an emerging topic at developer events and meetups are being held for the language. There is increasing talk that upcoming parts of the Linux kernel (such as drivers) will be developed in Rust. One of the infamous advertisements is a piece of malware written in Rust (detected this year) – it appears that even criminals value reliable code. Could we be witnessing the emergence of the successor of C/C++?
We must say that this project is developing rapidly. 2022 has seen support for new platforms, including Renesas’ RA series microcontrollers and WebAssembly, perhaps to your surprise. On top of that, the huge value of this issue can be seen from the traffic on the repository, i.e. patches and functionality uploaded. In addition, the community came together around this solution. All in all, the visible commitment and strong development of Micropython encourages (and will increasingly encourage) the use of this environment.
Artificial intelligence – this field is constantly evolving. Manufacturers are refining their frameworks so that embedded device developers, without having to study the intricacies of neural networks, can harness the benefits of this technology. Aside from this, AI itself, as well as the supporting hardware, is becoming increasingly powerful. So it’s no surprise, then, that it is already easy to find industry-specific solutions supporting 4K image analysis at 60 frames per second. There are also some exciting projects emerging where AI has been applied in areas that no one would have thought of at first. For example, using AI to reduce the burden on developers to meet the MISRA standard!
In simple terms, artificial intelligence is tasked with analyzing the deficiencies found in code. It determines which submissions should be ignored and which are worthy of the development team’s attention. A defect’s severity is also determined by the experience of the development team member addressing it. In this way, the team is less burdened with analyzing the code for quality. Isn’t that surprising?
We all know that robots are starting to become more and more commonplace in the world around us. Recent years have seen a remarkable boom in this industry – and I am not just referring to industrial robots, but even humanoid ones. Experts point out that the automotive industry and the education sector are driving the humanoid robot market. There are also a lot of upcoming companies emerging that want to adapt robots for the aerospace or military sectors. New developments in this field intrigue me greatly. During 2023, we can expect a number of interesting developments, not only as a result of the trends that have emerged over the past year but also as a result of a lack of components influenced by the next level of AI development or the search for programming languages that are more efficient and safer. Hardware manufacturers have announced many highlights. As for me, I’m very curious to see how using the RUST language in Linux will affect the stability of the operating system and its resistance to hacking attacks.
The topic of cybersecurity is something I’ll be keeping an eye on, as updating legacy solutions with more security features will keep many developers up at night, especially since slip-ups continue to occur, even in industries such as the automotive one.
