会员体验
专利管家(专利管理)
工作空间(专利管理)
风险监控(情报监控)
数据分析(专利分析)
侵权分析(诉讼无效)
联系我们
交流群
官方交流:
QQ群: 891211   
微信请扫码    >>>
现在联系顾问~
热词
    • 2. 发明公开
    • APPLIANCE FAILURE PREDICTION
    • EP3419223A1
    • 2018-12-26
    • EP18178431.5
    • 2018-06-19
    • THOMSON LICENSING
    • ERIKSSON, Brian CharlesPUDHIYAVEETIL, AjithSUN, YifanDING, Weicong
    • H04L12/28
    • G01M99/005H04L12/2823
    • A method to determine an operational status of one appliance includes sampling a plurality of sensors monitoring operation of a plurality of appliances. Sensor sampling may relate to energy input, temperature output, audio noise output, and vibration output. The sampling producing composite sensor measurements; the sensors not dedicated to any particular appliance. Filtering 215a, 215b, ... 215n the composite sensor measurements for information for the one appliance is accomplished to remove sampled data that represents appliances other than the one appliance. A comparison of the sensor data for the one appliance with the reference information of the one appliance is determined. If an operation anomaly is present in the information for the particular appliance, a notification is transmitted 240 to a user device concerning the operation of the one appliance. Sensor filtering blocks 215a, 215b, ...215n are filters used to discriminate for a specific device/appliance. This filter block is used to filter out other devices/appliances that might be operating concurrently in the environment. Sensor signal deviation blocks 220a, 220b, through 220n calculate the deviation, if any, from the filtered sensor data for a given device/appliance from block 215a, 215b, through 215n and the expected normal sensor data for one device/appliance from the database 210. The entire set of sensor deviation data outputs from blocks 220a, 220b, through 220n is a collection of deviations for one appliance. Block 230 calculates the sum of all sensor deviation calculations from the outputs of blocks 220a, 220b, through 220n. The output of the 220a, 220b, through 220n blocks can be considered multiple votes on whether there is a failure. The configuration is useful to collect, filter, and disambiguate sensor measurements so that appliance data may be analyzed to perform failure prediction and provide alerts to a user.