Method and system for activity recognition and behaviour analysis

Patent metadata

Field	Value
Publication number	US-2019377916-A1
Application number	US-201916434597-A
Country	US
Kind code	A1
Filing date	Jun 7, 2019
Priority date	Jun 8, 2018
Publication date	Dec 12, 2019
Grant date	—

Abstract

Energy remains a critical challenge for continuous sensing: with low-capacity batteries, wearable devices require frequent charging. In contrast, installing sensors in everyday ‘smart objects’, such as kitchen cabinets, household appliances and office equipment, supports ADL detection via indirect observations on human interaction with such objects, but cannot provide individual-specific insights in multi-tenanted environments. The embodiments herein provide a method and system for energy efficient activity recognition and behavior analysis. Architecture disclosed utilizes a hybrid mode of inexpensive, battery-free sensing of physical activities performed by a subject been monitored during his Activities for Daily Living (ADLs). The sensing combines object interaction sensing with person-specific wearable sensing to recognize individual activities in smart spaces. The method and system disclosed quantifies a probabilistic approach that uses longitudinal observations of user-item interactions, over each individual episode, to compute the anomalous behavior of the subject.

First claim

What is claimed is:

1 . A processor implemented method for activity recognition and behavior analysis, the method comprising:
seamlessly sensing, via one or more hardware processors using a Radio Frequency Identification (RFID) reader operating in a low power mode, sensor data received from at least one battery less primary sensor associated with at least one primary object ( 302 ), wherein usage of the at least one primary object is mandatory to initiate a primary activity associated with an ADL among a plurality of Activities of Daily Living (ADLs); detecting, via the one or more hardware processors, mobility of a subject to initiate the primary activity based on variation in the sensor data received from the at least one battery less primary sensor associated with the at least one primary object ( 304 ), wherein the variation in the sensor data indicates proximity of the subject with the at least one primary object; triggering, via the one or more hardware processors, the RFID reader to switch from the low power mode to a high power mode on detection of mobility of the subject for the primary activity ( 306 ), wherein the RFID reader in the high power mode receives RF data from:
at least one battery less wearable tag worn by the subject, and wherein the at least one battery less wearable tag comprises a RF powered passive accelerometer; and
a plurality of passive RFID tags tagged to a plurality of secondary objects placed in proximity to the at least one primary object, wherein one or more secondary objects from a plurality of secondary objects, required to be used by the subject while performing one or more secondary activities associated with each ADL among the plurality of ADLs to perform each ADL without presence of anomaly, are predefined;
analyzing, via the one or more hardware processors, the RF data received from the at least one battery less wearable tag over a plurality of window intervals to tag the primary activity of the subject, for each window interval, to an ADL among the plurality of ADLs ( 308 ); analyzing, via the one or more hardware processors, RF data received from the plurality of passive RFID tags of interest over each window interval to detect the one or more secondary activities performed by the subject using the one or more secondary objects and interaction of the subject with the one or more secondary objects corresponding to the tagged primary activity ( 310 ); and determining, via the one or more hardware processors, presence of anomaly in each of the window interval if the tagged primary activity of each window interval and the interaction of the subject with the one or more secondary objects in the detected one or more secondary activities for each window interval maps to a predefined primary object and secondary object usage criteria, wherein the predefined criteria is set for confirming expected execution of the ADL among the plurality of ADLs ( 312 ).

2 . The method of claim 1 , further comprising switching back, by the one or more hardware processors, the RFID reader to the low power mode when the variation in the sensor data received from the at least one battery less primary sensor indicates that the subject is not in proximity with the at least one primary object ( 314 ).

3 . The method of claim 1 , further comprising performing episode analysis, by the one or more hardware processors, by analyzing the received RF data, received from the at least one battery less wearable tag worn by the subject and the plurality of passive RFID tags tagged to a plurality of secondary objects, over a predefined time span comprising a plurality of window intervals to determine performance of the subject in performing the ADL over longer time interval ( 316 ).

4 . The method of claim 1 , further comprising generating, by the one or more hardware processors, an alert notification on detection of presence of the anomaly during execution of the ADL for corresponding window interval.

5 . A system ( 102 ), comprising:
a memory ( 202 ) storing instructions; one or more Input/Output (I/O) interfaces ( 206 ); and one or more hardware processors ( 204 ) coupled to the memory ( 202 ) via the one or more I/O interfaces ( 206 ), wherein the one or more hardware processors ( 204 ) are configured by the instructions to:
seamlessly sense via Radio Frequency Identification (RFID) reader ( 104 ) operating in a low power mode, sensor data received from at least one battery less primary sensor ( 106 ) associated with at least one primary object, wherein usage of the at least one primary object is mandatory to initiate a primary activity associated with an ADL among a plurality of Activities of Daily Living (ADLs);
detect mobility of a subject to initiate the primary activity based on variation in the sensor data received from the at least one battery less primary sensor ( 106 ) associated with the at least one primary object, wherein the variation in the sensor data indicates proximity of the subject with the at least one primary object;
trigger the RFID reader ( 104 ) to switch from the low power mode to a high power mode on detection of mobility of the subject for the primary activity, wherein the RFID reader in the high power mode receives RF data from:
at least one battery less wearable tag ( 108 ) worn by the subject, wherein the at least one battery less wearable tag comprises a RF powered passive accelerometer; and
a plurality of passive RFID tags ( 110 ) tagged to a plurality of secondary objects placed in proximity to the at least one primary object, wherein one or more secondary objects from a plurality of secondary objects, required to be used by the subject while performing one or more secondary activities associated with each ADL among the plurality of ADLs to perform each ADL without presence of anomaly, are predefined;

analyze the RF data received from the at least one battery less wearable tag ( 108 ) over a plurality of window intervals to tag the primary activity of the subject, for each window interval, to an ADL among the plurality of ADLs;
analyzing, via the one or more hardware processors the RF data received from the plurality of passive RFID tags ( 110 ) of interest over each window interval to detect the one or more secondary activities performed by the subject using the one or more secondary objects and interaction of the subject with the one or more secondary objects corresponding to the tagged primary activity; and
determining, via the one or more hardware processors, presence of anomaly in each of the window interval if the tagged primary activity of each window interval and the interaction of the subject with the one or more secondary objects in the detected one or more secondary activities for each window interval maps to a predefined primary object and secondary object usage criteria, wherein the predefined criteria is set for confirming expected execution of the ADL among the plurality of ADLs.

6 . The system ( 102 ) of claim 5 , wherein the one or more hardware processor ( 204 ) is further configured to switch back the RFID reader to the low power mode when the variation in the sensor data received from the at least one battery less primary sensor ( 106 ) indicates that the subject is not in proximity with the at least one primary object ( 314 ).

7 . The system ( 102 ) of claim 5 , wherein the one or more hardware processors ( 204 ) are further configured to perform episode analysis by analyzing the received RF data over a predefined time span comprising a plurality of window intervals to determine performance of the subject in performing the ADL over longer time interval.

8 . The system ( 102 ) of claim 5 , wherein the one or more hardware processors ( 204 ) are further configured to generate an

Assignees

• Tata Consultancy Services Ltd

Inventors

• Jaiswal Dibyanshu
• Gigie Andrew
• Ghose Avik
• Chakravarty Tapas
• Misra Archan

Classifications

• G16H50/20

  for computer-aided diagnosis, e.g. based on medical expert systems · CPC title

• A61B5/1125

  Grasping motions of hands · CPC title

• A61B5/1127

  using markers · CPC title

• G06K19/0723

  the record carrier comprising an arrangement for non-contact communication, e.g. wireless communication circuits on transponder cards, non-contact smart cards or RFIDs · CPC title

• G06K7/10128

  the step consisting of detection of the presence of one or more record carriers in the vicinity of the interrogation device · CPC title