Category Archives: iot

A Battery-less Future?

(Corrected: Atmosic has not (yet) won the GSA Award, but is short listed for it. The winner will be announced in December. Apologies)

At what point can we ditch batteries, the last encumbrance to our wireless nirvana?

The biggest single block on a wireless, connected future where everything everywhere is attached to chips and sensors which relay, receive and act on instructions from afar is power. And that means either that the device is connected to the electricity grid (which probably means you don’t need it to be wirelessly connected) or it has a battery in it. Which will need charging or replacing.

Long-range low-power technologies like low-powered wide-area networks (LPWANs think LoRa, NB-IOT and SigFox) have gone some way to solving this problem — instead of a power-hungry 4G modem you have a simple chip that sends only the most necessary data and runs off a battery that can run for years — but that doesn’t solve the problem of more complex or power-hungry devices that need to communicate more frequently and more loquaciously. These endpoints will need someone to service them. Internet of Things, Interrupted.

But what if the devices could find their own energy? What if they could “scavenge the energy they need to operate from whatever naturally occurring electrons were in their environment, regardless of that environment”, in the words of Chris Rust, founder and general partner of VC investor Clear Ventures?

Energy harvesting, as it’s called, is not new. Solar power is in effect harvesting the sun’s rays and turning it into energy via photovoltaic cells; wind or wave turbines do something similar (called electrodynamics). But scavenging ambient energy in the immediate environment into electrical power will yield only a few watts at most — enough to augment batteries or, possibly, to replace them. (Still enough to power your solar calculator indoors, and solar power is highly efficient at conversion.)

Energy harvesting can be done an in a number of ways:

  • kinetic energy — vibrations, stress, tension or movement using piezoelectric materials, for example. Imagine the vibration on aircraft wings being converted to energy, or the reverberation of heartbeats to power a pacemaker. (Some examples of vibration energy harvesting can be found here from ReVibe Energy of Gothenburg.)
  • Other examples of vibration-based energy harvesters are triboelectric charging — when certain materials are separated one becomes electrically charged (think the static electricity from running a comb through one’s hair) or the more traditional electromagnetic vibration, where relative motion between magnet and coil induces current into the coil. (Think turning a door knob or hitting a switch.)
  • Then there’s temperature — where differences across a thermoelectric crystal cause a voltage, or the temperature of a pyroelectric crystal changes, generating a charge. The new PowerWatch, for example, uses both thermoelectric — the heat emitted from your wrist — and solar charging. The device uses chips from Matrix Industries.)
  • Then there’s radio frequency (RF) radiation, emitted from routers and cell towers, or from RF chargers, that transmit electromagnatic waves in a specific area. So while this might be scavenging in the sense that it is capturing wasted or existing radiation, it could be deliberate — say, via pointing an RF source at your remote device and switching it on.

So some of this is happening. A RFID (radio-frequency identification) or NFC (near field communication) sticker (think price tags) or chip (think contactless cards, or has no battery in it, instead harvesting the power from the device connecting to it through a technique called backscatter, which transmits data by reflecting modulated wireless signals off a tag and back to the reader.

In the labs of academia the vision is that the body becomes a patchwork of, well, patches, where the energy is derived from the body itself to power unobtrusive sensors which monitor our health: solar-powered heart sensors no bigger or less flexible than a Band-Aid, or sensors that draw their power from the natural conductive properties of skin, storing their energy in stretchable capacitors made of carbon nanotube forests (so called because the material grows like trees 30 micrometers tall, their canopies tangled on wafers.)

But for now, the movement is in industry, and buildings. Companies like EnOcean sells self-powered switches and sensors for maintenance-free lighting which draw their power either, in the case of switches, from the kinetic movement of being pressed or in the case of sensors, from light (indoor and outdoor) or temperature differences to detect occupancy, say.

The changes will really kick in when devices can generate enough energy to be able to transmit over significant distances wirelessly. That means WiFi, which requires a decent-sized battery, rather than, say, Bluetooth, which has too short a range to be any use beyond your headset, mouse or keyboard. That, however, may not be true for much longer. The latest version of Bluetooth, version 5, expands its range by four times, making it comparable to WiFi. And companies like Atmosic Technologies believe they can extend a Bluetooth device’s battery life by between 5 times, to, well, forever.

Atmosic Technologies, just announced as winner of shortlisted for the Global Semiconductor Association’s startup of the year, says that “with the advent of Bluetooth 5, combined with ultra-low-power functionality, power consumption is low enough to be supported by harvested RF, light, or heat energy, while still able to provide the range and coverage equivalent to Wi-Fi.” In short, it makes “the concepts of “forever-battery” and “battery-free” IoT realistic. IoT devices can work for the lifetime of the devices on the batteries they come with, or without batteries at all.”

Atmosic says its a fully integrated single chip with RF energy harvesting (for size see the image at the top of this post) can provide small form factor battery-free operation up to a distance of several meters from the RF source. This could be a game changer, because it would mean not only that all your Bluetooth devices would not require charging, but that they could communicate over longer distances. It would also mean a lot more devices could communicate with each other without you having to worry about whether they need charging. But Atmosic acknowledges that “this is the first step in the journey,” which sounds as if we’re still some ways off the battery-free IoT revolution.

The Phantom Prospects of 5G

Source: Light Reading
Source: Light Reading

Telcos are once again touting a new generation of mobile protocols as their saviour. And once again, we should raise an arched eyebrow. 

A great piece by Mike Dano from Light Reading dissects the reasons behind — and the challenges facing — telcos’ push for 5G. And why you shouldn’t be expecting 5G to rock your world any time soon.

The reasons are relatively simple. The above chart from Wall Street analysts MoffettNathanson show how, for US telcos (and the story is similar elsewhere) there’s a lot more data traffic (the black line) , but a lot less money being made per user (the blue line).

So telcos need something new. Hence 5G. But how are they going to make money out of it? The old thinking about 4G was that the improved speeds and bandwidth was going to enable telcos to make more money out of each user. But that hasn’t happened (see chart again.) So why do they think 5G will be any different?

5G is going to be very expensive to deploy. This is not just a question of attaching a new box or upgrading an existing one to an existing basestation or tower. Mike quotes MoffettNathanson in a recent research note on Verizon. “Deploying 5G networks with enough density to meet expectations for blazing fast 5G speeds will cost real money. If not from charging more for using more, then where will the money come from?”

Atop that are the problems of heat, which I’ve written about before.

So the most likely outcome of 5G, at least in the first few years, is that it’s aimed at business. Think the Internet of Things (IoT): connected (and self-driving) cars, smart cities, healthcare, industrial automation. Or outdoor surveillance cameras, according to a Gartner report cited by Mike, which says this is likely to be the largest market for 5G in the short term.

But even then, there are questions about this. This won’t be cheap for enterprises. And 3G and 4G hardly shone when it came to connecting devices. Attaching a modem to a device and then having that ping back is expensive — installing and replacing a battery, if necessary, having and managing a SIM card for each device, etc.

Ericsson, one of the cheerleaders for mobile IoT, acknowledges in a June report that despite its grand claims for a mobile IoT future, most mobile IoT devices are actually using 2G and 3G:

Today, the majority of cellular IoT devices are connected via 2G and 3G technologies (GPRS, EDGE and HSPA). The number of legacy connections is expected to increase slightly until 2022, and then remain stable throughout the rest of the forecast period.

Hard to imagine that’s a huge source of revenue.

And then there’s narrowband IoT, which I’ve written about before. Yes, 5G offers the kind of speed and low latency that would be attractive for a lot of use cases, but when you just need to send a few bytes narrowband is much more appealing. And needn’t involve the telcos at all.

Most likely outcome? Telcos are going to have to demonstrate, somehow, to customers that the new use cases they’re hoping will save them will actually work. And that means working closely with them or investing in them, or vice versa. Build it and they will come might be the mantra, but that is going to require a lot of faith that they will indeed come, and in a timeframe that makes sense. New verticals don’t pop up overnight.