Low-Power Bed / Seat Occupancy Sensor Based on EMFi - Foundations on Natural and Artificial Computation - page 502

The sensor network consists of a set of sensor nodes, to be distributed ubiq-
uitously at home. The sensor nodes implement the minimum architecture to
carry out environmental monitoring activities, develop messages with informa-
tion about events that occurred and to form a wireless network with mesh topol-
ogy over which the messages are taken to the PC. These nodes comprise a
microcontroller and a radio transceiver specifically for low-power, as well as a
set of commercial sensor devices able to pick up the activity at home in a non-
intrusive and safe way. In particular, passive infra-red (PIR) sensors are used to
monitor the location of the user, pressure-detecting mats under the mattress to
detect the presence in bed and magnetic contact sensors to signal the opening
of the front door of the house.
2.2 Bed/Seat Occupancy Detector Issues
Pressure mats have been appropriate as a first approach to detect the bed or seat
occupancy. However, their functionality is dependent on various environmental
factors such as the type of mattress and bed base or weight of the user. It's
operation is simple. It consists of two foils separated by a layer of foam with
cylindrical holes. Both sensor foils conform a normally-open contact (NO) that
closes under the weight of the user. The main disadvantage of this device is
that, once made, its sensitivity is determined by the diameter of the holes. The
sensitivity required is dependent on the type of mattress, bed base and weight
of the user. The use of mats with different sensitivity to each individual is not
feasible on a large scale. For these reasons finding adaptive systems to maintain
consumption and bounded cost it is desirable.
The use of sensors resistive based transducers could be an alternative, such as
FlexiForce [7] from Tekscan Manufacturer. It consists of two sheets of polyester
substrate. The transducer is modelled as a variable resistance dependant of pres-
sure on the sheet. Researchers at Massey University in New Zealand have developed
a bed occupancy sensor which generates a digital signal indicative of the presence
by implementing a configurable threshold [5], [6]. This device may be suitable for
AAL systems. However, the authors do not refer to the energy consumption of the
implemented sensor and a quick inspection of the proposed circuit shows that con-
sumption will always be greater than 2.5 mA. This makes this design is inappro-
priate for a network of wireless sensors powered by batteries, with a global average
consumption (including sensing, data processing and communications) below 3mA.
A third alternative is proposed in this paper, based on MEMS ( Micro-Electro-
Mechanical Systems ). Specifically, EMFi ( Electro-Mechanical Film [8]) based
traducers can be used to implement flat capacitors whose capacity is dependant
of the force exerted over it's surface. These devices, properly covered, may be
placed under one leg of the bed to allow a rough measure of the exerted force.
Sensor node would apply periodic signals filtered by a circuit tuned by the EMFi
device and, allowing capacity and, therefore, the exerted weight measuring. The
sensor node would be calibrated on place, before their operation. This solves the
problem of lack of adaptability. Re-calibration routines could be also made using
information from the context extraction and reasoning layers.