Last night, I was scrolling through my LinkedIn feed and came across a VDC survey report on embedded engineers that said, “A substantial portion of engineers do not see upside revenue potential from the IoT over the long run.” Yikes!
To wit, or lack thereof, my response was, “Designers at the bench see through the IoT hype, true, but that's mostly because the benefits and upside to IoT don't accrue to them, it accrues to the owners of the data.”
Sometimes you wish you could just hit rewind and edit yourself. It’s not that the comment was wrong, as such. If you know the embedded space and the engineers within it, you’d appreciate their extremely sensitive hyperbole antenna (aka: B.S. detector, to use the vernacular.) They’re a hard-core bunch, which is why we have such great embedded systems.
So what’s an embedded system? Even embedded system designers, developers, and industry insiders have a hard time defining it. In short, embedded systems are anything besides a PC, laptop, notebook or tablet that does computing. Now you see why even embedded systems designers don’t call themselves embedded systems designers and have a hard time identifying with the term. Instead they’re data server designers, or automotive electronics designers, or smart coffee makers developers.
(Data analysis, one of the fundamental aspects of the IoT, has enabled the development of low-cost, reliable ICs and modules like Intel’s new XMM 6255M modem, pictured right. For embedded systems, the IoT is already lowering costs and improving system designs to optimize the business.)
Identity crisis aside, embedded systems have historically been standalone, and didn’t really connect to anything but a power outlet. Think: vending machine. Then vending machines got GPRS cellular data capability and today even they are Internet connected.
So why the skepticism? Even the VDC report indicated that IoT adoption is at an all-time high in a market that has traditionally favored isolation. “51% of surveyed engineers are currently working on a project that features Internet connectivity and cloud support,” said the report. “Furthermore, more than three-quarters of surveyed engineers reported that their organizations are either already deploying IoT-capable devices and applications, or plan to deploy them by 2020.”
A hint as to what fuels the skepticism then points to one of the most fascinating aspects of the IoT. “…but engineers tend to see through much of it [the hype], as they are intimately familiar with IoT products,” said the report. This intimacy comes from putting together the ingredients of an embedded system: electronic sensors, analog signal chains, processors, memory, power supplies, I/O and wired or wireless connectivity elements, and then slinging code to make it all function reliably. The fundamental principles, theory and skills required to do this design don’t change, just the requirements. In the case of IoT, those requirements include ever-lower power consumption, more sensors, and reliable connectivity.
At first blush, it seems clear that the benefits of IoT bypass the embedded developer at the bench, who gets stuck with the increasingly demanding requirements. And so off my comment went.
But something about that didn’t sit right with me, even as I sent it. Later, just as I was dozing off to sleep, the reason for my comment anxiety occurred to me: Optimal+!
Bear with me a second: Optimal+ has spent years gathering data on semiconductors as they are manufactured, to the point that they can reliably tell which wafers and die will be of the highest quality. Their clients can then charge a premium for that ultra-reliable IC for designers of systems for applications such as automotive, industrial, military and aerospace.
That process of gathering and analyzing data toward a better outcome is classic IoT, and embedded developers are benefitting from it directly, through lower-cost ICs that push the boundaries of performance, power and reliability so they can meet their own design requirements.
It’s this kind of analysis, done internally at top tier semiconductor companies, that has made possible the likes of Intel’s Atom, Edison and Galileo, with some new devices announced just last month for automotive, industrial, and smart-building communications and processing.
But for embedded developers, the upside is just beginning. In speaking with David Park over at Optimal+, it quickly became clear that the next step is to extend the analysis up the supply chain so that system designers and integrators will know immediately which parts to put together for optimum performance and reliability.
However, what Park didn’t discuss is the next level, which is the application of sensors and communications within the embedded system to relay back to the manufacturer the real-time state of the system and analyze its failure modes, if any, so the system can be improved.
MEMS sensors can indicate falls, temperature sensors can show the ambient temperature, and low-cost global navigation satellite system (GNSS) parts can show location of incidents. All this can be fed back to the system designer, and to the manufacturing plant, to improve the design and make an embedded developer’s life a bit easier. There’s nothing quite like real-time feedback to help monitor a system’s health and affirm a good design.
So, if I could go back in time, I’d change my comment to something like, “The embedded developer or engineer could be forgiven for not seeing the long-term value of the IoT for their business given they are immersed in the details of building it out. However, not only is the IoT making their own task easier and creating more demand for their capabilities, but it’s also creating the opportunity to innovate more rapidly as they incorporate IoT technologies and techniques into their own design flow.”
Maybe it’s good that LinkedIn comments has a word limit.