Energy Harvesting – The lights may be going out for ZigBee…

Every farmer knows that if they want a good harvest they need to take care where they sow their seed.  One of the first principles they’ll learn is to sow the seeds on their own land, not their neighbours.  So I was somewhat shocked to see the recent announcement from the ZigBee Alliance about their new Energy Harvesting profile, dubbed ZigBee Green Power.  In their press release they talk about a feature set to establish a global, standard technology for self-powered devices operating through energy harvesting techniques.

There’s nothing fundamentally wrong with that as a goal.  Energy harvesting’s a fascinating area of technology, which has only recently emerged from research into reality, as better generating technology is devised, along with lower power radios.  It’s taken around twenty five years to come to fruition, during which companies and researchers have been actively patenting their ideas and techniques.  Those patents don’t just cover the energy harvesting devices, they cover all of the parts of the chain – generating the power, converting it to a form that can be used, storing it, connecting to a radio and transmitting information.  They also cover the applications, such as wireless light switches.

Hence my amazement at the naivety of the ZigBee Green Power press release.  Whether or not ZigBee can come up with a specification that is able to run on a few tens of microJoules of power is irrelevant – I’m sure they can as they’ve bright people working in the specification group.  What is far more important is whether or not it will be legal for anyone to ship a device that is based on it, as it will almost certainly infringe the Intellectual Property of the main stakeholders.  So the press release looks like either an ill thought out, opportunistic attempt to regain some momentum, or a worrying piece of evidence that the ZigBee Alliance has lost the plot… 

Scavenging energy, whether it’s vibrational, thermal or mechanical has been a holy grail for several decades.  It’s particularly attractive in building automation, where sensors and switches that run off energy harvesters can be installed without wiring.  The biggest potential market is for light switches, closely followed by thermostats.  For that reason it’s attracted the big names in this industry, including Siemens (which group formed into Enocean), Honeywell and Emerson.  Today Enocean is the front runner in terms of products and real deployments, as well as in having a very impressive patent portfolio.

A lot of players in the market are involved in just one part of the equation.  They might be making a specific energy harvesting device, such as Perpetuum, a particularly low power radio, like Nordic Semiconductor,  or an optimised battery, like Infinite Power Solutions. They’re each busy generating IP and patents, but it typically only covers their particular part of the equation.

Which takes us back to Enocean.  They’ve done the same thing of generating specific IP, but the amount of time they’ve been playing means that they’ve gone much further and patented complete applications.  Take one of their key patents – WO98/36395 (an english version of which is granted in the US as US 7,230,532).  It covers (my descriptions in italics): “A sensor system comprising: at least one voltage generator for conversion of non-electrical energy to electrical energy (that’s the energy harvesting device), at least one energy store coupled to the voltage generator (typically a capacitor or battery), a processor controller, at least one voltage converter having an input coupled to the energy store  and an output coupled to said processor controller (a charge controller), at least one sensor (which could be a light switch) coupled to said processor controller, at least one wire free transmitter (a radio) coupled to an output of said processor controller for wire-free transmission of transmission messages from said processor controller which contain at least one measured value from the at least one sensor, wherein a timer circuit is provided, to be triggered as a function of a voltage level of the at least one energy store to activate the sensor system, a specific time interval after being triggered, in order to transmit at least one transmission message”. 

For those who aren’t familiar with reading patent claims, this is a pretty comprehensive description of a device which harvests energy, stores it locally with some electronics to optimise the storage regime, then at the detection of a state change or switch input, uses that stored energy to turn on a radio and transmit a message.  The bit about the timer ensures that the sensor has enough power to give a reliable reading. 

This is where the ZigBee announcement is so naïve.  All that ZigBee can specify is their radio.  They may be able to get around any existing low power radio IP, but as soon as anyone attempts to deploy it in a working capacity, they’ll come up against the current patent coverage.  The example of a light switch that ZigBee gave at the conference that accompanied their press release appears to me to blatantly infringe Enocean’s patent.  Yet the gung-ho promotion and press announcement contained no warning to potential implementers that their product development could be fruitless.

There are lots of good things about ZigBee, but an announcement like this is not one of them.  Developers need to understand patent issues – you can’t just jump on a bandwagon with impunity.  My experience of Enocean is that they’re a good, solid company that has worked hard to get to their leading market position.  They’ve set up the Enocean Alliance to allow others to get access to their technology.  If a newcomer wants to play, the route should be to work with the IP holders, not trample over them. 

Some of the companies supporting this latest ZigBee initiative should know better, as they’re already involved in associated markets.  Maybe it’s time for their patent lawyers to give a few lessons to the respective development teams.  And for ZigBee to look a little harder at each new bandwagon it intends to jump on.  If it really has this little understanding of IP, then its own standard needs to be read with a very large caveat emptor.