I recently connected with the CEO of a fascinating company based in Cork, Ireland, called Treemetrics. Its mission is to eliminate waste in the logging industry by moving beyond antiquated forest survey methods and other problems caused by a lack of good information. By applying data analytics technology to this sector, the company claims to be able to reduce wastage by up to 20%.
To make this work, Treemetrics has to extract data from exotic sensing systems such as 3D laser mapping to obtain stem volume and taper information, which tells a lot about the size, type and health of tree groupings. It’s pretty cool and leaps beyond where the forestry industry has been for literally hundreds of years in terms of harvesting efficiency.
Seeing the forest for the trees
Most intriguing about what Treemetrics and many other firms are doing is the potential for cross-correlation between sensor data sets that at first glance don’t appear to have anything to do with each other.
For example, automakers are starting to consider the use of vehicle-generated data streams to inform external constituents. Given that the typical modern car has to sample outside air temperature and barometric pressure in order to efficiently control fuel mixtures, that data coupled with geolocation could produce highly accurate meteorological maps that provide pinpoint information in near real time.
Similarly, seeing the instances where vehicles suddenly come to a stop or swerve might alert a traffic management operator or image analyzer to look at a camera in that location, to determine if a pothole has formed or if some obstacle is blocking the roadway. That data is there, as it’s required to operate traction control and anti-lock braking systems. It’s just a matter of anonymizing it and sending it to the right recipients.
Your phone is a sensor, naturally
Another interesting cross-correlation is the counting of mobile devices as a proxy for the number of humans in a given locale, and their movements. It doesn’t require your GPS to be turned on; merely the fact of being on the mobile network allows your carrier to triangulate on the device’s location with reasonable accuracy.
This data can be used to determine where and at what times people gather, such as around a bus stop, or when they’ve gotten into cars or boarded trains by seeing the acceleration of their phones (cool, eh?). Knowing how many people are transiting through a given area might, for example, inform a coffee shop chain as to the ideal place for its next store.
The laws of nature meet the law of Moore
What’s really happening here is a combination of Moore’s Law in action (i.e., better, smaller, faster, cheaper processors) and the evolution of sensors themselves.
I recently came across a tiny multi-axis gyro that can output movement in a wide G-force range with amazing accuracy. What made it truly amazing was the ultra-low pricing for even modest quantities. It turns out that it’s one of the sensors contained within a hugely popular game controller. The manufacturer had to provide essentially military-level accuracy and durability in a device that would be produced in the many millions. This made it wonderfully inexpensive.
So one can now imagine building it into a sensor package that is cheap and very small, to affix to migrating animals (the Internet of Caribou?). Migration patterns could be tracked with incredible accuracy, and inform environmental experts along with commercial enterprises that might be operating along their pathways. This even applies to the skies, where air traffic might be temporarily routed around bird migrations. Helpful, since not all flocks show up well on radar.
In agriculture, very low-cost moisture and temperature sensors can now be applied to entire crop regions, some in the ground, some on the plants themselves, even on swarms of drones.
‘Star Trek’ gives us a bright glimpse
I always liked it when the Starship Enterprise pulled up to a new planet and the bridge officers would give the captain a thorough rundown of what the ship’s sensors told them about the constituency of the planet, including the atmosphere, level of societal and technological evolution, surface temperature and more. The ability to quickly gather sensor-derived data was depicted as essential (all in service of the plot, of course).
We’re pretty much there, in terms of sensing capabilities, sans the cool starship. Increasingly and on a planetary basis, there will be sensors embedded in, or attached to, many objects and living things, including plants, animals and ourselves.
How can organizations profit from billions of sensors?
As with most technology adoption cycles, the processes of normalizing, publishing, locating and obtaining sensor information will undergo “friction reduction” over time (think of Amazon, Bloomberg, eBay and Uber as examples of friction reduction leading to booming growth in usage). This means that your organization in most cases won’t have to install and own the sensors in order to derive benefit from them. Your key opportunities will be in innovative uses for the sensor data streams that will become increasingly available.
It’s conceivable that for certain industries and civilian/governmental agencies, sensor experts will join data scientists as being vital to the evolution of operating paradigms. As with the emergence of the Internet and mobile communications, the big data coming from large groupings of sensors and processing of same represent an evolution in our usage of technology to massively affect business operations and consumer lives. The possible uses for physical-world sensor data are endless, and as the Internet of Trees example shows, we’ve barely scratched the bark, er, surface.
This article is published as part of the IDG Contributor Network. Want to Join?