All of this came to the forefront for us on a recent visit to Cluster Wireless. Cluster is based in San Diego and is in early stages of solving two of the problems we identified above. Namely: how to connect to carrier networks and how to make for easier endpoint programming.

Cluster has taken a new approach to this. Or not so much a new approach as a modern approach. They offer a software layer that should work with any merchant baseband. That layer abstracts all the messiness of embedded processor programming. Users can then program their M2M module using common programming languages. Their kernel is relatively small, so the software can fit on the M2M module without any extra components or memory. This means it is cheaper than using an Android OS which theoretically could fill the same role. Cluster’s software is also presumably much more secure than Android. Cluster’s solution is the middle ground between unwieldy embedded programming and some other, heavier solution. Cluster could the operating system (OS) for M2M.

Why does this matter?

In the examples above, we noted that M2M solutions need to be fitted to a variety of end user functions. Cluster’s solution lets users program their own features. For instance, some users will want to get alerts from their M2M module only at times of a major alert – say a burglar alarm going off or a dilithium crystal overheating; others will want to receive data on a regularly scheduled basis, say a power meter or pipeline volume gauge. Cluster lets all these users buy common radio components, and then customize to fit.

Cluster is privately-held, and still in very early days, but we think they have the potential to really alter the M2M landscape.