2028: Open source is out, app-centric chips are in - Embedded.com

2028: Open source is out, app-centric chips are in

November 2028 is the 40th anniversary of ESD. Click here read other 2028 lookbacks.

Every new year, I find myself reminiscing about the past. This time, as I was going through my box of “artifacts,” I found the iPhone I bought in 2008. It was a beautiful piece of engineering for its time, and in fact, a product that forever changed the way people looked at embedded systems. I remember that the iPhone and I were inseparable. It went where I went because it gave me access to almost every piece of information I wanted, instantly. At the time, you had to type everything using one or maybe two fingers. A slow process, but we didn't mind. The most absurd thing at the time was text messaging. Imagine this; you conversed with someone by typing one character at a time, he or she replied, and you had to read the response!

I could not imagine going back to those days. Everything is so much easier now with our current embedded devices. As an example, embedded devices have changed the medical industry in huge ways. I remember visiting my physician once or twice a year and having to wait sometimes hours in a “waiting room” before he could see me. In fact, when it was finally my turn, I went in another “private” waiting room where a nurse took my vitals and sometime afterwards, I was finally seen by my physician who was always in a hurry and didn't have time to get to know me. Thanks to the embedded devices I have in my skin (most look like small tattoos), I only have to visit the physician when my embedded computer sees something that needs his expertise.

In 2008, we powered our cars with fuel refined from crude oil. During rush hour, millions of cars all around the world wasted an incredible amount of fuel just “sitting” in traffic, not to mention the tons of greenhouse gases they produced. With today's high-efficiency solar-powered vehicles, at least vehicles are no longer contributing to global warming. Traffic jams have nearly been eliminated by the thousands of sensors and computers in our vehicles as well as those placed on the roadways. Now, computers efficiently move vehicles along their desired destination.

After aggressively pursuing open-source software in the late '00s, the industry realized that it was too costly a choice, with some products failing and developers discovering that they had no resources to fall back on for support when the software didn't perform as required. This led the embedded industry to rapidly create and adopt stringent coding standards for software, effectively halting the pursuit of open source. I was quite fortunate to be one of those championing the need for embedded software standards. And myself, along with some of my colleagues, were instrumental in making this happen as we had been recognized as producing high-reliability software for years.

Having delivered embedded software to the industry now for over 30 years, I did not, and I don't think anyone could have, anticipated the prominence software would come to hold in embedded systems, nor the fact that software costs would far exceed hardware costs. Chips are now fully application-centric (as opposed to the old, generic chips), due in large part to the advances in embedded software to support the increased performance and complexity of modern processors. The shift to processors as a commodity with software determining their functionality, applications, and performance, grew out of the development of software standards.

The product failures that pushed the embedded industry to adopt standards for software development actually benefited the industry to a huge degree as it wasn't long before the complexity of software increased such that it could no longer be written one line at a time. With the increase in discipline around software development and the commoditization of RTOSes, the industry saw a shift to the use of visual tools that facilitated code reuse. Because of the growing complexity of embedded software, these visual tools were implemented to quickly put “software components” together that could automatically select and configure chips; this eroded the importance of CPU architectures and instruction sets, de-emphasizing the chips themselves. Interesting and unanticipated outcomes of these developments were the obstacles that developed in debugging systems; historic approaches were no longer effective. Instead, chips started to be designed for self-verification, and visibility had to be built into the targets, further justifying the need for and adoption of software standards.

What was truly impressive, however, was less what was done in the world of embedded systems, but rather what embedded systems made possible in the “real world.” In 2008, the world was facing a number of growing crises–energy production, pollution and other environmental damage, and shrinking natural resources. As was hoped and anticipated, it was technology that enabled the global community to address and ultimately overcome these issues. Back then, who could have imagined the evolution of voice recognition or integration of speech and vision into electronic devices.

We live now in a world of solar and electric cars, where we all telecommute and computing systems offer truly seamless “virtual offices.” 3-D video conferencing has become the norm, significantly reducing global reliance on air travel. We're seeing the benefits of a significant investment in wind, solar, and tidal farms that have achieved truly clean electricity. And this magazine itself is possibly one of the most visible changes, being offered exclusively in an electronic format on tablet PCs that have been adopted globally with their touch interfaces and voice inputs to facilitate use while increasing efficiency and productivity. The Internet has undergone more transformations than one can count but has become the exclusive outlet for news, movies, music, and communications.

It's humbling to think of the number of embedded systems that operate in the modern world; even more humbling is all of the time and effort that went into their development. These systems have brought about and enabled the very necessary change that took place in one of the shortest technological transition periods in modern history.

Jean J. Labrosse is president of Micrium. He is a regular speaker at the Embedded Systems Conferences. Jean is the author of two books: MicroC/OS-II, The Real-Time Kernel and Embedded Systems Building Blocks, Complete and Ready-to-Use Modules in C and has written numerous articles for magazines.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.