WO2018158106A1

WO2018158106A1 - Multi-mode sensor

Info

Publication number: WO2018158106A1
Application number: PCT/EP2018/054097
Authority: WO
Inventors: Leendert Teunis Rozendaal; Martinus Wilhelmus BLUM; Daniel Martin Goergen
Original assignee: Philips Lighting Holding BV
Current assignee: Signify Holding BV
Priority date: 2017-03-02
Filing date: 2018-02-20
Publication date: 2018-09-07
Anticipated expiration: 2019-09-02

Abstract

A presence sensing device comprising: a presence sensor configured to generate an output signal indicative of sensed presence; signal processing logic configured to simultaneously apply to the output signal first presence detection processing to identify a first signal characteristic in the output signal and second presence detection processing to identify a second signal characteristic in the output signal; wherein: the first presence detection processing is configured to generate a first message of a first message type in response to identifying the first signal characteristic in the output signal; and the second presence detection processing is configured to generate a second message of a second message type different from the first in response to identifying the second signal characteristic in the output signal.

Description

MULTI-MODE SENSOR

TECHNICAL FIELD

The present disclosure relates to presence detection within an environment.

BACKGROUND

Presence sensors are used for a wide variety of purposes in which data from the sensor is analyzed to trigger events, for example to trigger a change in a lighting effect rendered by a lighting system (e.g. to turn on in response to presence detection) or to trigger an alarm in a building (e.g. in response to motion detection). Many types of presence sensors are known in the art which may detect presence, e.g. of a human user, directly (e.g. thermal imaging cameras) or indirectly (e.g. via a location of a user device such as a smartphone), or may detect motion within an environment (e.g. Passive Infrared sensors).

In the field of motion (and general presence) sensing, typically filtering is applied to the raw signal from the sensor before deciding if action needs to be taken (such as switching on the light or triggering an intruder alarm). The signal from the sensor can also contain noise or spikes which should not trigger the light or the alarm. In a sensor design, a trade-off needs to be made between (fast) reaction times and likelihood of false triggers. Typically, filtering makes the decision more reliable but inherently takes detection time (latency) and/or reduces the sensitivity of the sensor.

EP0838792A2 discloses a multifunction passive infrared occupancy sensor which functions as an occupancy sensor for security systems and also as an occupancy sensor for energy management control systems.

SUMMARY

To illustrate the kind of trade-offs that might need to be considered in existing sensor designs, consider the following. A presence sensor for triggering a luminaire of a lighting system to turn on in general just needs to have a fast response time, as a user is generally less concerned with possible false positives arising due to sensor noise and more with responsiveness. On the other hand, a user of a security system in which the presence sensor triggers an alarm (e.g. an intruder alarm) will want to minimize false positives, and is less concerned about a fast response time. Hence, sensors for lighting systems tend to apply little to no filtering to the sensor signal before generating triggers, whereas sensors for security systems do apply filtering to the signal before generating a trigger e.g. of an alarm. Lighting and security are just examples and it is understood that the amount of filtering used depends on the implementation.

As such, currently available presence (e.g. motion) sensors are either optimized for short latency (e.g. lighting use cases) by applying no or limited filtering, but accepting some false triggers, or optimized for low false trigger rate (e.g. soft-security use cases) by applying more filtering, but accepting longer latency. If a sensor allows adjustment of the filtering parameters, it might be possible to configure it to function for either type of use cases, but not at the same time.

Note that the number of false positives (and quickness of detection) can be affected by various factors during processing of an output signal from a sensor. For example, a positive "presence detection" may be determined when the strength of the signal rises above a threshold - lowering this threshold may result in a quicker detection which is more prone to false positives (due to e.g. noise in the signal).

The inventors have recognized that it is desirable to have a single presence sensor unit which can simultaneously operate in both modes simultaneously, providing both types of trigger.

An example situations where both use cases are wanted from a single sensor is a presence sensor in a home: if the user is at home, he wants the sensor to control the lighting so it has to react quickly; if he is not at home, he wants the same sensor to inform him (e.g. via a text message) if someone is detected. In the security case, latency is of no concern but the user does not want false triggers.

The present invention makes it possible to use a single sensor for multiple use cases simultaneously, including but not limited to security and lighting.

Hence, according to a first aspect disclosed herein, there is provided a presence sensing device comprising: a presence sensor configured to generate an output signal indicative of sensed presence; signal processing logic configured to simultaneously apply to the output signal first presence detection processing to identify a first signal characteristic in the output signal and second presence detection processing to identify a second signal characteristic in the output signal; wherein: the first presence detection processing is configured to generate a first message of a first message type in response to identifying the first signal characteristic in the output signal; and the second presence detection processing is configured to generate a second message of a second message type different from the first in response to identifying the second signal characteristic in the output signal.

That is, a single presence sensing device applies two types of presence detection to the same sensor signal in parallel. The first and second presence detection processing are independent of each other in that they can arrive at different conclusions on presence detection due to the fact that they look to identify different characteristics in the output signal.

In embodiments, the first presence detection processing has a first response time and the first signal characteristic is the output signal indicating presence which persists for at least a time interval determined by the first response time; and the second presence detection processing has a second response time and the second signal characteristic is the output signal indicating presence which persists for at least a second time interval determined by the second response time and exceeding the first time interval.

In embodiments, the second presence detection processing comprises applying filtering to the sensor signal and the second response time is determined by at least one parameter of the applied filtering.

In embodiments, the first message comprises an indicator of the first message type and the second message comprises an indicator of the second message type.

In embodiments, the indicators of the message type are network endpoint identifiers.

In embodiments, the indicators of the message type are identifiers of a respective destination of the first and second messages.

In embodiments, the first presence detection processing has a first threshold signal amplitude and the first signal characteristic is the output signal exceeding the first threshold signal amplitude; and the second presence detection processing has a second threshold signal amplitude greater than the first and the second signal characteristic is the output signal exceeding the second threshold amplitude.

In embodiments, the first signal characteristic is a first frequency of presence indications in the output signal; and the second signal characteristic is a second frequency of presence indications on the output signal different from the first.

In embodiments, the first signal characteristic is a first rise time of the output signal; and the second signal characteristic is a second rise time of the output signal different from the first. According to another aspect disclosed herein, there is provided a presence sensing device comprising: a presence sensor configured to generate a sensor signal for detecting presence; a first processing module configured to apply to the sensor signal first presence detection processing (i.e. presence detection processing of a first type), having a first response time, so as to generate a first message comprising an indicator of detected presence and an indicator of the first response time, the first message generated in response to changes in the sensor signal occurring over a first time interval determined by the first response time; and a second processing module configured to apply to the sensor signal second presence detection processing (i.e. presence detection processing of a second type) simultaneously with the first signal processing, having a second response time greater than the first response time, so as to generate a second message comprising an indicator of detected presence and an indicator of the second response time, the second message generated in response to changes in the sensor signal occurring over a second time interval determined by the second response time and exceeding the first time interval.

That is, a single presence sensing device applies two types of presence detection to the same sensor signal in parallel. The first and second presence detection processing are independent of each other in that they can arrive at different conclusions on presence detection due to their different response times. The indicator of the response time in each message provides a recipient of that message with information on how the conclusion conveyed by that message (i.e. its presence indicator) has been arrived at.

For example, each message can comprise information identifying the type of that message (indicator off message type).

Note the indicator need not convey an actual value for the response time (though that is not excluded) - it is sufficient for the indicators to simply distinguish the two processes with the different response times. For example, it can be a network endpoint ID, where the two processes are associated with different respective endpoint IDs. That is, the indicator of a response time in each message is an indicator of which one of the types of presence detection has been applied to generate the indicator of detected presence in that message, thereby indicating the response time of that type of processing.

In embodiments, the indicator of the first response time is a first network endpoint identifier of the presence sensing device associated with the first processing module, and the indicator of the second response time is a second network endpoint identifier of the presence sensing device associated with the second processing module.

In embodiments, the network endpoint identifiers are ZigBee network endpoint identifiers.

In embodiments, each indicator of detected presence is a binary value indicating that presence has been detected.

In embodiments, the presence sensor is a motion sensor.

In embodiments, the indicators of the first and second response times comprise first and second confidence values respectively, which increase with the first and second response times respectively, the second confidence value greater than the first.

According to a second aspect disclosed herein, there is provided a method implemented at a presence sensing device and comprising steps of: receiving an output signal indicative of sensed presence from a presence sensor; simultaneously applying, at the presence sensing device, to the output signal first presence detection processing to identify a first signal characteristic in the output signal and second presence detection processing to identify a second signal characteristic in the output signal; wherein: the first presence detection processing is configured to generate a first message of a first message type in response to identifying the first signal characteristic in the output signal; and the second presence detection processing is configured to generate a second message of a second message type different from the first in response to identifying the second signal

characteristic in the output signal.

According to a third aspect disclosed herein, there is provided a computer program product comprising computer executable code embodied on a computer-readable storage medium configured so as when executed by one or more processing units to perform the steps of: receiving an output signal indicative of sensed presence from a presence sensor; simultaneously applying to the output signal first presence detection processing to identify a first signal characteristic in the output signal and second presence detection processing to identify a second signal characteristic in the output signal; wherein: the first presence detection processing is configured to generate a first message of a first message type in response to identifying the first signal characteristic in the output signal; and the second presence detection processing is configured to generate a second message of a second message type different from the first in response to identifying the second signal

characteristic in the output signal. BRIEF DESCRIPTION OF THE DRAWINGS

To assist understanding of the present disclosure and to show how embodiments may be put into effect, reference is made by way of example to the

accompanying drawings in which:

Fig. 1 shows a presence sensing device according to the present invention;

Fig. 2 shows an example ZigBee implementation;

Figs. 3A and 3B illustrate data processing in accordance with an embodiment of the present invention; and

Figs. 4A and 4B illustrate data processing in accordance with another embodiment of the present invention.

DETAILED DESCRIPTION OF EMBODIMENTS

The present invention allows a single sensor to be used in both "lighting" and "soft security" uses cases. A single physical sensor device presents two "logical sensors" towards the system network (e.g. towards: a controller, or central controller, of a lighting network, sometimes called a bridge of a lighting network; ZigBee light; or the Internet). Each logical sensor is optimized for a particular use case:

"lighting": sensor reaction time needs to be fast; some false triggers are acceptable

"soft security": sensor needs to be reliable; false triggers are not acceptable but latency is not a real issue

The different behavior for the two logical sensors can be implemented in the physical device by either:

applying two software algorithms in parallel to the raw input signal of the motion sensor, one optimized for lighting use cases, one optimized for soft security use cases, or

applying a single algorithm with two decision output signals (one for lighting use cases, one for soft security use cases).

Generally, embodiments of the present invention relates to generating two distinct messages from a single sensor output signal by simultaneously applying a respective two processing algorithms. Each algorithm is configured to "look for" a different

characteristic of the signal and to generate its message when that characteristic is identified. The sensor output signal can generally be viewed as a presence value (analogue or digital) which varies over time (either continuously or discontinuously) and hence the characteristics may be a variety of properties of this signal. For example, if the characteristic is a time for which the signal is above a threshold, then the first processing may generate its message in response to the signal persisting above the threshold for a first time and the second processing may generate its message in response to the (note: same) signal persisting above the threshold for a second, longer, time. Alternatively or additionally, the thresholds for each of the first and second processing may be different.

Figure 1 shows a presence sensing device 100 comprising a presence sensor 101, a processor 103, and communications interface 105. The processor 103 is operatively coupled to each of the presence sensor 101 and the communications interface 105, e.g. via a wired connection.

The presence sensor 101 comprises one or more sensing units arranged to detect a presence within an environment. For example, a Passive Infrared (PIR) sensor as is known in the art. In any case, the presence sensor 101 generates a data stream of a sensed value within the environment, which is discussed in more detail below. The presence sensor 101 provides the data stream to the processor 103.

The processor 103 comprises one or more processing units such as CPU(s) running code so as to perform the functionality described herein. The processor 103 is shown in Figure 1 as executing a first processing module 113 and a second processing module 123 which represent functionality of said code. The processor 103 is configured to apply processing to the sensor data received from the sensor 101 in order to generate messages, and to provide the generated messages to the communications interface 105.

Alternatively, some or all of this functionality could be implemented in dedicated hardware such as one or more Field-Programmable Gate Arrays (FPGAs) or Application Specific Integrated Circuits (ASICs). That is, in general a processing module can be implemented in software, hardware, or any combination thereof. Particularly, it is noted that the first and second processing modules may be implemented as a single processing module configured to perform the functionalities of both as described herein separately for the purposes of explanation only.

The communications interface 105 comprises one or more wired or wireless communication means (ports) configured to at least transmit data to an external device. The communications interface 105 may also be configured to receive data from the or another external device. For example, the communications interface 105 may comprise a wireless interface configured according to the ZigBee, WiFi, or Bluetooth wireless standard, and be configured to transmit the messages received from the processor 103 to an external device accordingly. The following description is given in terms of ZigBee wireless communications, but it is understood that it applies equally to other wireless (and wired) communication means known in the art.

In embodiments in which the communications interface 105 comprises two or more wired or wireless communication means, messages generated by the first and second processing modules 113, 123 may be transmitted via different respective communication means. Additionally, the communications interface 105 may present two "virtual" interfaces or endpoints to the external device (i.e. to the network to which the device 100

communicates, in which case the messages may be transmitted via different virtual interfaces or endpoints. The communications interface 105 is shown in Figure 1 as having a first endpoint 115 and a second endpoint 125, which it is understood represent either individual ports of the interface (distinguished in the physical layer), or virtual endpoints of the same physical interface (which are distinguished in the network layer).

As a short summary, the first processing module 113 is arranged to receive the sensor signal from the sensor 101 and to apply a first presence detection processing algorithm to generate a first message to be transmitted via the interface 105. This algorithm has a low latency (a quick response time), for example by generating the message immediately when the sensor signal exceeds a threshold value (as soon as a non-zero presence is detected).

Note that the "response time" may generally refer to how quickly the algorithm determines, the output signal of the sensor, that presence has been detected. This may be due to an explicit variable in the algorithm which determines a data collection time (e.g. a time window over which the sensor data is integrated), or may be due to other factors, some of which are mentioned below. What matters is that a first signal processing (applied by the first processing module 113) and a second signal processing (applied by the second processing module 123) look for different characteristics of in the sensor output signal. This may be a time-window over which the signal is integrated (or averaged, or for which the signal must persist e.g. above a threshold value) but may also be one or more of the following:

A duration of the signal. E.g. if the sensor output signal indicates that the presence detection lasts less than or more than (depending on context) a threshold amount of time (e.g. 50 milliseconds) a message of a first type is generated. The threshold amount of time for the second message generation may be different.

An amplitude of the signal. E.g. if the amount of movement detected indicated by the sensor signal is less than a predetermined amount of movement, a message of the first type is generated. The threshold amount of movement for the second message generation may be different. This relates to the first and second processing having different amplitude thresholds when analyzing the output signal for what they consider to be a "presence". E.g. the first processing may consider a 50% signal strength to be indicative of presence, whereas the second processing may require 90% before it considers a presence to have been detected.

A frequency of the signal(s). E.g. if fewer than a predetermined number of signals (e.g. fewer than five) within a predetermined time period (e.g. ten seconds) then a first message is generated. The number of signals and/or the predetermined time period required to generate the second message may be different.

A rise time or rise time pattern of the signal. The rise time of a signal is a known term describing the amount of time taken for the signal to change from one state to another. E.g. in this case the first processing could look for a first rise time value in the output signal being the amount of time taken for the signal to change from "no presence" to "presence". For a continuously- varying output signal (e.g. where 0 is "no presence" and 1 is "presence", and the signal may be anywhere in between), the gradient or average gradient of the signal over time when approaching 1 can be used.

The second processing module 123 is also arranged to receive the sensor signal from the sensor 101, and applies a second presence detection processing algorithm to generate a second message to be transmitted via the interface 105. The algorithm in this case, as opposed to the first, has a higher latency (a longer response time than the first). This algorithm does not simply generate the message immediately in response to the sensor signal exceeding the threshold, like the first, but "waits" for a predetermined amount of time following this and only generates the message if the presence is still detected in the signal for that amount of time. The threshold values of the first and second processing modules given in the above examples may be different or may be the same.

The two messages are transmitted via the interface 105 and each comprise an indicator of the detected presence and an indicator of the response time of the respective algorithm which was used to generate it. The indicator of the response time generally refers to any information which allows direct or indirect determination of whether the message originates from the first processing module or the second processing module. In some embodiments this response time indicator is an explicit value indicating the response time, in other embodiments it is a network endpoint identifier of the endpoint used to transmit the message (e.g. an indicator of the communication port, or virtual interface). In any case, the extraction of the first message from the sensor data by the first processing module 113 may be considered a first "logical sensor" and the extraction of the second message from the sensor data by the second processing module 123 may be considered a second "logical sensor" in the sense that they each generate different output messages, but there is only one physical sensor (sensor 101) actually detecting a property within the environment.

Note that the response time (or indeed other signal characteristic, as mentioned above) indicator does not have to be an identifier of the virtual endpoint (i.e. the source of the message) and may instead by an identifier of the address to which the message is sent (i.e. the destination of the message), which may be an IP address. In these cases, the messages are sent to different destination addresses and hence the characteristic used (which processing algorithm generated the message) is implicitly known by each receiving device. For example, a system could be arranged in which the first message (and subsequent other messages generated by the first processing algorithm) is always sent to a first device and the second message (and subsequent other messages generated by the second processing algorithm) is always sent to a second device.

The information from the two logical sensors described may be exposed on multiple ZigBee endpoints using the standard ZigBee mechanism.

In ZigBee, "clusters" define the low-level protocol syntax for groups of commands and attributes. Each cluster is identified by a hexadecimal identifier (ID). A device supporting a cluster is able to perform all tasks defined by that cluster, such as reporting and auctioning commands. For example, cluster 0x0006 defines on/off commands for basic switches as 0x0006,0x00 (off), 0x0006,0x01 (on), 0x0006,0x02(toggle). In this example, the Occupancy Cluster (0x0000) and Occupancy Sensor Type Cluster (0x0001) are harnessed without any modification to the existing ZigBee protocol.

An example of such an arrangement is shown in Figure 2, in which a different respective endpoint (endpoint 201 and endpoint 202) is provided for each logical sensor. Each endpoint has an "Occupancy Sensing" cluster with the "normal" messaging (cluster attribute "Occupancy" with Id=0x0000 reporting 0x01 for "occupancy detected" and 0x00 for "non-occupancy detected"). Note that each of these clusters could have the standard (and/or manufacturer-specific) attributes to set e.g. sensitivity. This allows an existing 'consumer' (i.e. recipient) of such ZigBee messages (like a luminaire or bridge of a lighting network) to use the information from one (or both) of these endpoints since the information is presented using the standard ZigBee mechanism for Occupancy Sensing.

The different endpoints have the same network address, but are identified by their different endpoint IDs (0x0000 and 0x0001 in this example). A device accessing the endpoints has to choose between two identical looking endpoints, which can be thought of as having a manufacturer specific "capabilities" attribute which distinguishes the different functions.

Figures 3A and 3B show an embodiment in which two endpoints (e.g. the first endpoints 115 and second endpoint 125 shown in Figure 1) are used. These endpoints are referred to as EP1 and EP2 respectively (as shown in Figures 3 A and 3B).

ZigBee Attribute Report messages are labelled "AR". In the present invention, the attribute is a presence within the environment and the report can be either AR(1) indicating that a presence is detected within the environment, or AR(0) indicating the end of a presence which was earlier detected (i.e. that no presence is detected). That is, an AR(1) message to an external device informs the external device that the sensor 101 has detected presence within the environment, and this may be followed (although it is optional) by an AR(0) message at a later point informing the external device that the presence is no longer sensed in the environment.

In Figures 3 A and 3B, both EP1 and EP2 are configured to transmit AR(1) and

AR(0), though it is still appreciated that AR(0) is optional. The difference between the configuration is the response times at which the endpoints operate. EP1 is configured according to a short response time and EP2 is configured according to a longer response time.

Consider the sensor data 301 shown in Figure 3 A which exhibits a characteristic 311 indicative of presence detection. This characteristic 311 can be identified as a deviation in the sensor data 301 from a background level. For example, the sensor data 301 may be zero when no presence is detected and non-zero when presence is detected, or may be a constant background level when no presence is detected and an elevated level which may fluctuate when presence is detected (the latter being shown in Figure 3 A). EP1 represents the output of messages generated by the first processing module 113 via endpoint 115 and EP2 represent the output of messages generated by the second processing module 123 via endpoint 125.

As is shown in Figure 3 A, the algorithm used by the first processing module 113 generates the message substantially immediately in response to the detection

characteristic 311 which is shown as AR(1) sent be EP1. For example, this may be in response to the sensed value being outside of a background range. Hence, it is understood that EP1 provides a rapid response (low latency) but may be susceptible to false positives because, for example, the sensed value may randomly fluctuate to outside the background range. In the example of Figure 3 A, the characteristic 311 is a long (or strong) presence detection - the second processing module 123 also processes, in parallel to the first processing module 113, the sensor data 301 and generates another attribute report. Unlike EP1, the attribute report by EP2 is not sent "straight away". Instead, the second processing module 123 applies more filtering to the sensor data 301 than the first processing module 113 (NB the first processing module 113 may apply no filtering). The application of filtering to sensor data is, per se, known in the art, and results in a more "confident" presence detection in that background noise is less likely to cause a false positive, but at the cost of increased latency.

Hence, AR(1) sent by EP2 is sent at a later time than AR(1) sent by EP1, as is shown in Figure 3 A. That is, the first processing module 113 and second processing module 123 each process the same data (sensor data 301) and each generate attribute reports indicating that presence has been detected in the environment, but which differ in the latency and confidence levels. An external device receiving one or both of these attribute reports and therefore able to determine which version of the attribute report is which based on the endpoint from which it was received. E.g. the external device can tell whether an attribute report is the low-latency but low-confidence one or the high-confidence but high-latency one based on the endpoint network address. Therefore, the external device can choose which attribute report to use based on the requirements of the particular implementation. E.g. if the external device is a lighting system, it might choose to use the low-latency presence indication (attribute report from EP1) and ignore the other attribute report.

Based on the above, it is appreciated that the "AR(1)" event on EP1 can be sent briefly after the start of the event since little filtering (with little latency) is needed. On the other hand, the "AR(1)" event on EP2 is sent later since more filtering (with more latency) is needed to increase the level of confidence that someone is actually present as needed for "soft security" use case, e.g. to wait for several full cycles of the PIR signaling in Figure 3A.

Note also that Figure 3 A shows AR(0) messages being sent by each endpoint. These indicate to the external device that the presence characteristic 311 has ended, and are sent after a "hold time" after the end of the characteristic 311 (e.g. when the sensor data 301 has returned to the background level). The hold times for both use cases might be different, hence the "unoccupied" messages "AR(0)" on EP1 and EP2 may be sent at different times.

Figure 3B shows the same logical sensors in an example in which a short presence characteristic 312 is present in sensor data 302. This presence characteristic 312 comprises the sensor data 302 being outside the background range, which results in EP1 sending its AR(1) message, but for only a short amount of time, which means that the sensor data 302 has returned to the background level before EP2 generates its AR(1) message. In other words, the characteristic 312 is long enough for EP1 to generate a message, but not EP2. This can be understood conceptually as EP1 having a sensitive "view" of the sensor data 302 which is able to "see" short term data fluctuations such as characteristic 312, but EP2 having only a more coarse view in which it only response (can only "see") fluctuations which persist for an amount of time. This amount of time is dependent on the amount of filtering applied to the sensor data 302 by the second processing module 123, as is known in the art.

The above-outlined mechanism can be extended beyond two endpoints when more than two levels of detection (or levels of "confidence") are introduced. In these cases, a third endpoint may be provided for transmitting messages generated by a third processing module having a response time longer still than the second processing module 123.

The examples outlined above in relation to Figures 3 A and 3B concern implementations using standard ZigBee Occupancy endpoints. An alternative is to combine the messages for the two (or more) detection levels and two or (more) endpoints onto a single ZigBee endpoint, with different messages for the various use cases (e.g. lighting and soft- security). In this case, a new cluster can be defined with extra information (beyond the occupied/unoccupied bit from the standard ZigBee Occupancy cluster). An example of this is shown in Figures 4A and 4B which mirror the sensor data shown in Figures 3A and 3B (which are hence given the same reference numerals) but differ in the implementation of the processing and messages, as outlined below.

In the Figure 4A, a longer or stronger motion event is detected, leading to messages for both use cases. This is analogous to Figure 3 A, but all messages are sent from a single endpoint EP and include explicit information pertaining to which processing module was used to generate the message. That is, instead of the external device receiving the messages being able to tell which latency level is a given message was generated using based on the endpoint providing the message, it is able to tell based on information included in the message itself (application layer data).

In this example, the first processing module 113 is indicated by a "1" and the second processing module 123 is indicated by a "2". This is, "AR(1)1" is an attribute report that presence detection has occurred as determined by the first processing module 113 and "AR(1)2" is an attribute report that presence detection has occurred as determined by the second processing module 123. These may be referred to as "detection level 1" and

"detection level 2", respectively.

Note that the "AR(1)1" event (detection level 1) can be sent briefly after the start of the event since little filtering (with little latency) is needed. On the other hand, the "AR1(2)" event (detection level 2) is sent later since more filtering (with more latency) is needed to increase the level of confidence that someone is present as needed for "soft security" use case, e.g. to wait for several full cycles of the PIR signal.

In Figure 4B, as in Figure 3B, a short or limited presence event is detected. In this case, only "AR(1)1" and "AR(0)1" are transmitted to the external device as the presence does not persist long enough to trigger AR( 1 )2.

Again, the hold times for both use cases might be different, hence the

"unoccupied" messages "AR(0)1" and "AR(0)2" may be sent at different times.

The single network endPoint providing both response time outputs, as in Figures 4A and 4B, may be provided by a standard ZigBee Occupancy Sensing cluster, along with some manufacturer-specific cluster on the same EndPoint. That is, in the description above there were two Standard Zigbee Clusters, but it is also possible for one to be manufacturer specific and the other standard. One of these clusters is then used for one of the detection levels (e.g. lighting use case) while the other cluster is used for the other detection level (e.g. soft-security use case). A standard "consumer" of messages will be able to use the sensor for one of the detection levels. A manufacturer-specific "consumer" will be able to use the sensor for both detection levels. It is possible for one (e.g. AS(x)l) to be a default to be used by any receiving device, and the other (e.g. AR(x)2) to be used only be certain (e.g. particular permitted) devices and to be "hidden" from the other devices which are not permitted to use the second detection level.

Standard Zigbee clusters are standardized applications. This means that they can be used by other (e.g. third party) devices implementing this standard interface right away (without requiring any further configuration than that of the "factory" setting). Hence, when two Occupancy clusters are used for the different detector functionalities described above, the third party device can choose between one of the two virtual sensors (or both) without understanding the difference of behavior.

If a cluster is modified to be manufacturer specific, on the other hand (resp. the Zigbee standard is extended), the difference in behavior is made explicit via an API and appears as such to the third party device (e.g. a user's phone), e.g. the same but with additional (manufacturer specific ) API extension when the two functions are combined into one. As this is not standardized, only applications understanding these manufacturer specific extensions can make use of it.

Another variation of the above is where on a single EndPoint (e.g. the lighting function) the standard ZigBee Occupancy Sensing cluster is exposed, and extending this cluster with one (or more) manufacturer-specific attributes (e.g. to provide the security function). The standard "Occupancy" attribute of the cluster is then used for one of the detection levels while the manufacturer-specific attribute is used for the other detection level. A standard "consumer" of messages will be able to use the sensor for one of the detection levels (the one reported using the standard "Occupancy" attribute). A manufacturer- specific "consumer" will be able to use the sensor for both detection levels.

The embodiments described above in which one virtual sensor is standard and the other is manufacturer specific have the advantage that the sensor has a fall back mode for the case when the sensor is used in combination with systems that cannot distinguish between the two use cases (as this system will only use the normal occupancy cluster or attribute, and ignore the separate cluster or attribute). To take advantage of this fall back mode, the sensor should use the normal occupancy cluster for its main use case (either lighting or soft security).

It will be appreciated that the above embodiments have been described only by way of example. Other variations to the disclosed embodiments can be understood and effected by those skilled in the art in practicing the claimed invention, from a study of the drawings, the disclosure, and the appended claims.

For example, although the above has been described in terms of a "light" use case and a "security" use case, both the output presence detection indications could be used by the same system to trigger two different behaviors of the same external device or system. E.g. the different levels of "certainty" of the events to both control lighting use cases - the initial presence indication (quick but possibly some false positive) will give some light and the second presence indication (slower but hardly any false positives) will give full light.

This can be extended further by sending multiple (different) AR(l)events, e.g. at 50%, 80% and 90%> confidence levels. This allows the application receiving the sensor events to choose the sensitivity it needs for a specific use case (e.g. low sensitivity for switching lights, medium sensitivity for switching other devices (e.g. turn on music), high sensitivity for sending notification, highest for sending warnings,...). These may all be output by the same interface where one of more functions (e.g. lighting function) is standard and one or more other functions (e.g. soft security) are manufacturer specific, as mentioned above. At this point it is noted explicitly that an increased response time (longer latency) generally means a higher confidence, and that means of converting between the two schemas are known in the art. For example, it is possible to determine, mathematically, the length of the response time necessary in order for the confidence in the detection to be X%, where X is an arbitrary "target" confidence.

In embodiments in which the multimode sensor of the present invention provides multiple levels of sensitivity/certainty, the filtering method (either in the device itself, or in the bridge/controller receiving the signals) can be adapted based on the 'state' of the home, which can be based on a location of a user, e.g. differentiate between 'home/away', and/or a time of day, e.g. 'day/night'.

In the claims, the word "comprising" does not exclude other elements or steps, and the indefinite article "a" or "an" does not exclude a plurality. A single processor or other unit may fulfil the functions of several items recited in the claims. The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measures cannot be used to advantage. A computer program may be stored and/or distributed on a suitable medium, such as an optical storage medium or a solid- state medium supplied together with or as part of other hardware, but may also be distributed in other forms, such as via the Internet or other wired or wireless telecommunication systems. Any reference signs in the claims should not be construed as limiting the scope.

Claims

CLAIMS:

1. A presence sensing device comprising:

a presence sensor configured to generate an output signal indicative of sensed presence;

signal processing logic configured to simultaneously apply to the output signal first presence detection processing to identify a first signal characteristic in the output signal and second presence detection processing to identify a second signal characteristic in the output signal; wherein:

the first presence detection processing is configured to generate a first message of a first message type in response to identifying the first signal characteristic in the output signal; and

the second presence detection processing is configured to generate a second message of a second message type different from the first in response to identifying the second signal characteristic in the output signal;

wherein the first message comprises an indicator of the first message type and the second message comprises an indicator of the second message type and wherein the indicators of the message type are network endpoint identifiers.

2. The presence sensing device of claim 1, wherein: the first presence detection processing has a first response time and the first signal characteristic is the output signal indicating presence which persists for at least a time interval determined by the first response time; and the second presence detection processing has a second response time and the second signal characteristic is the output signal indicating presence which persists for at least a second time interval determined by the second response time and exceeding the first time interval.

3. The presence sensing device of claim 1, wherein the second presence detection processing comprises applying filtering to the sensor signal and a second response time is determined by at least one parameter of the applied filtering.

4. The presence sensing device of claim 1, wherein the indicators of the message type further are identifiers of a respective destination of the first and second messages.

5. The presence sensing device of claim 1, wherein the network endpoint identifiers are ZigBee network endpoint identifiers.

6. The presence sensing device of any preceding claim, wherein each the first message of the first type and/or the second message of the second type comprises a binary value indicating that presence either has or has not been detected.

7. The presence sensing device of any preceding claim, wherein the presence sensor is a motion sensor.

8. The presence sensing device of claim claim 2, wherein indicators of the first and second response times comprise first and second confidence values respectively, which increase with the first and second response times respectively, the second confidence value greater than the first.

9. The presence sensing device of claim 1, wherein the first presence detection processing has a first threshold signal amplitude and the first signal characteristic is the output signal exceeding the first threshold signal amplitude; and the second presence detection processing has a second threshold signal amplitude greater than the first threshold signal amplitude and the second signal characteristic is the output signal exceeding the second threshold signal amplitude.

10. The presence sensing device of claim 1, wherein the first signal characteristic is a first frequency of presence indications in the output signal; and the second signal characteristic is a second frequency of presence indications on the output signal different from the first frequency.

11. The presence sensing device of claim 1 , wherein the first signal characteristic is a first rise time of the output signal; and the second signal characteristic is a second rise time of the output signal different from the first rise time of the output signal.

12. A method implemented at a presence sensing device and comprising steps of: receiving an output signal indicative of sensed presence from a presence sensor;

simultaneously applying, at the presence sensing device, to the output signal first presence detection processing to identify a first signal characteristic in the output signal and second presence detection processing to identify a second signal characteristic in the output signal; wherein:

13. A computer program product comprising computer executable code embodied on a computer-readable storage medium configured so as when executed by one or more processing units to perform the steps of:

receiving an output signal indicative of sensed presence from a presence sensor;

simultaneously applying to the output signal first presence detection processing to identify a first signal characteristic in the output signal and second presence detection processing to identify a second signal characteristic in the output signal; wherein: