It’s still hard to design an IoT device, especially one that will be used for an extended period of time. The odds are high that the application requirements will change, possibly requiring a hardware upgrade. However, thanks to always-on connectivity, hardware can be future-proofed, while also converting it to a portal for post-sale incremental revenue streams.
For any designer of small or large systems for the IoT, the trade-offs and optimizations depend on the application. If it is a large server, the focus tends to be on performance at almost any cost. If it is a small, remote, wireless node at the edge the focus shifts to low cost, low power, small size and ruggedness. For the latter, performance tends to be sacrificed as high-end ICs can be expensive. Or at least they used to be.
Thanks to the industry’s adherence to Moore’s Law, the cost of powerful semiconductors has fallen dramatically. It’s now at the point where semiconductor vendors have shifted to adding functionality, design services and technical support for a given footprint and cost, rather than simply slashing prices.
For providers of IoT solutions, this has three distinct benefits. The first is lower cost per feature, that’s obvious. Next is the benefit from faster time to market due to better support and easy-to-use modules and kits from semiconductor manufacturers.
The third benefit is the option to provide the functionality expected by the end customer at the time of sale, while offering more features and functionality over time. This possibility comes from enabling “dark” or “latent” hardware that costs almost nothing to include, but can provide valuable functionality to the customer, post sale (Figure 1).
Figure 1: Philips Hue app-controlled LED lighting is interesting and has obvious aesthetic appeal. As the cost of silicon falls, an LED bulb can also become a sensor hub, for everything from acoustics to temperature, humidity and vibration. (Image source: Philips)
The model of “unlocking” additional features or functionality post-sale is not new, of course. Software vendors, from Microsoft to Adobe have been doing it since the dawn of the Internet. In app stores, for-fee features are always being added to what were initially free apps.
In the electronics industry, the sea change was most evident in the context of expensive test and measurement equipment. Instead of buying a new $50,000 oscilloscope or other piece of equipment, vendors responded to customer demands by letting them pay for the functionality they needed initially, then enabling more features within the same box over time, as needs changed and applications became more demanding.
Initially, these features were enabled upon entering a code provided by the hardware supplier. Later, thanks to ubiquitous connectivity and the ability to perform secure firmware updates automatically, the features could be added and activated, post sale.
Future-proofing the IoT bottom line
For designers of IoT systems, future-proofing used to mean fretting over how much more hardware to add at extra cost in the vague hope that a device will be able to perform whatever functions the customer asks two years hence, without needing to do an expensive full-scale hardware and software upgrade. This meant over-specifying processors, memory and communications ports and sensing capabilities, while still managing to make a decent margin.
Now, as wired or wireless over-the-air (OTA) updates become commonplace and relatively secure, the cost of over-specifying the hardware instead turns into an opportunity to add value and revenue. For example, most ICs have multiple built-in temperature sensors that help monitor the IC silicon and keep it from overheating. One or more of those sensors can also be brought out and “activated” post sale on an IoT device that may initially have been sold as a light switch or connected LED lamp. Now an LED lamp becomes a temperature monitor for in-room climate control.
The advent of low-cost micro electromechanical systems (MEMS) has taken the IoT space by storm in the form of highly accurate, acoustic sensing and processing for Amazon’s Alexa and Google’s Home voice-controlled artificial intelligence assistants. Those same MEMS acoustic devices can be added at almost zero cost to any IoT system, and turned on at a later date once the software for a new feature that can use acoustic sensing on an IoT device has been developed, again post sale.
For providers of IoT solutions, the hard, up-front decisions with regard to future-proofing a design have been replaced by a new way of thinking that can make customers happier, while adding to the top – and bottom – line.