Until now the role of wireless in motion detection has been confined to communicating the information, but now it is ready to become the sensing medium as well. There are a few emerging consumer systems available, but the most significant development so far has come from Quantenna, which has incorporated wireless motion sensing in its WiFi chips.
This has a similar range of applications to conventional visual detection systems such as security cameras, but with some clear advantages. Among these are avoiding need for new hardware, because motion detection can be added as a software upgrade to wireless routers and repeaters already present in many homes and businesses. It also does not rely on line of sight so provides ubiquitous whole home coverage without having to install at least one camera, and often two, in every room.
Quantenna selected Aerial Technologies, a Montreal- based start-up founded in 2015 to supply the technology, which it says has undergone rigorous testing equating to years of continual operation in the field. This is a prime example of statistically based machine learning since it relies on the fact that every object and its movement generate unique patterns of signal variation that the system can be trained to recognize in a given configuration. Strength of signals as measured at a particular point is reduced and transformed according to the objects they encounter in transit and the latter’s velocity of motion. So like a basic security camera it can detect motion, but has no trouble distinguishing between, say, a human and a cat.
There is scope for incorporating wireless motion detection in pay TV UIs for functions like turning on a TV set when someone enters the house or a given room. There are also numerous potential use cases associated with smart home services, ranging from security to turning down heating or air conditioning when the last person leaves the house, to triggering an alarm if an elderly person has been in the bathroom longer than 15 minutes.
Quantenna and Aerial Technologies have been testing such applications over WiFi networks incorporating these chips based on the latter’s Aerialitix package. This comprises two components, a software agent in the chip and cloud-based machine learning software which can be applied to tailor services to specific end customers. The two companies say they are now working with Tier 1 ISPs on some of the specific use cases with a view to commercial deployments. These operators are hoping their end subscribers will be willing to pay for additional smart home services that really do add value and yet can be enabled without additional hardware or an engineer’s visit.
Radiomaze, founded in 2014, is another start-up specializing in wireless motion detection with a similar approach. This comprises a kit made of two TP-Link TL-MR3040 WiFi repeaters running OpenWrt with Radiomaze’s motion detection software embedded. OpenWrt is an open source embedded operating system based on Linux, with the advantage here being that it will run on many available WiFi repeaters irrespective of the chipset.
Given that WiFi motion detection can be turned on so readily, its success will depend on how strongly service providers promote it and whether it does turn out to be robust enough in the field.