US 11,994,935 B2
Apparatus and method for predicting remaining lifetime of environment data collection sensor in livestock house
Seng Kyoun Jo, Daejeon (KR); Se Han Kim, Daejeon (KR); You Jin Kim, Daejeon (KR); Dae Heon Park, Daejeon (KR); Hyeon Park, Daejeon (KR); Jee Sook Eun, Daejeon (KR); Jae Young Jung, Daejeon (KR); and Won Kyu Choi, Daejeon (KR)
Assigned to ELECTRONICS AND TELECOMMUNICATIONS RESEARCH INSTITUTE, Daejeon (KR)
Filed by ELECTRONICS AND TELECOMMUNICATIONS RESEARCH INSTITUTE, Daejeon (KR)
Filed on Feb. 10, 2022, as Appl. No. 17/669,124.
Claims priority of application No. 10-2021-0019172 (KR), filed on Feb. 10, 2021.
Prior Publication US 2022/0253350 A1, Aug. 11, 2022
Int. Cl. G06F 11/07 (2006.01); G06N 3/044 (2023.01); G06N 3/08 (2023.01); G06Q 50/02 (2012.01)
CPC G06F 11/0757 (2013.01) [G06F 11/0736 (2013.01); G06N 3/044 (2023.01); G06N 3/08 (2013.01); G06Q 50/02 (2013.01)] 9 Claims
OG exemplary drawing
 
1. A method of predicting a remaining lifetime of an environment data collection sensor installed within a livestock house, the method comprising:
collecting livestock house environment information from a sensor through a wired/wireless communication interface;
storing the collected livestock house environment information in a database;
detecting an error of the sensor from the collected livestock house environment information;
generating a dataset for learning based on given livestock house environment information collected during a period from a time of sensor installation up to a time of detecting the error, the given livestock house environment information being obtained from the collected livestock house environment information stored in the database;
generating a sensor lifetime prediction model to perform the learning, the sensor lifetime prediction model being a time series-based machine learning model;
training the sensor lifetime prediction model using the dataset; and
predicting a remaining lifetime of the sensor at a specific time point by applying a dataset, which is acquired from livestock house environment information collected at the specific time, to the trained sensor lifetime prediction model.