# US 12,216,686 B1 · Machine learning technologies for predicting results of cable fire tests > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,216,686 B1 - Document ID: US12216686B1 - Status: Granted - Kind: B1 - Application number: 18/417,490 - Filed: 2024-01-19 - Granted: 2025-02-04 - Earliest claimed: 2024-01-19 - Assignees: Ul LLC - Applicants: Ul LLC - Inventors: Christopher Hauman; Anthony Tassone; Jiyuan Kang - Web page: https://patentgrep.com/patent/US12216686B1 ## CPC - G06F 16/285: {Clustering or classification} - G06N 20/00: Machine learning - G06N 20/20: Ensemble learning - G06Q 30/018: Certifying business or products - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) - G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS - G06Q: INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR ## Abstract Systems and methods for using machine learning models to predict an outcome of a product test are described. According to certain aspects, an electronic device may calculate, based on a received set of small-scale results as a first input to a first machine learning model of a plurality of machine learning models, a first result predicting an outcome of the product tested according to the large-scale product test. The electronic device may then calculate, based on the set of small-scale results as a second input to at least one second machine learning model of the plurality of machine learning models, a second result predicting the outcome of the product tested according to the large-scale product test. The electronic device may then predict an outcome of the large-scale product test based at least on the first result and the second result. ## Claims ### Claim 1 A computer-implemented method for predicting an outcome of a large-scale product test, the computer-implemented method comprising: receiving, by one or more processors, a set of small-scale results of a product tested according to a small-scale product test representative of the large-scale product test; calculating, by the one or more processors and based on the set of small-scale results as a first input to a first machine learning model of a plurality of machine learning models, a first result predicting an outcome of the product tested according to the large-scale product test, wherein calculating the first result includes: determining, by the one or more processors, a first classification for the product, and calculating, by the one or more processors, a confidence value for the first classification; calculating, by the one or more processors and based on the set of small-scale results as a second input to at least one second machine learning model of the plurality of machine learning models, a second result predicting the outcome of the product tested according to the large-scale product test, wherein calculating the second result includes: determining, by the one or more processors and based on a test profile for the large-scale product test, a second classification for the product, wherein determining the second classification includes: predicting, by the one or more processors and based on the set of small-scale results as the second input to the at least one second machine learning model of the plurality of machine learning models, the test profile, wherein: the at least one second machine learning model includes a plurality of regression models, and each regression model of the plurality of regression models predicts a test value of a plurality of test values of the test profile, the plurality of test values including at least one of: (i) a flame spread value; (ii) a total heat release value; (iii) a peak heat release rate value; and (iv) a fire growth rate index value; and predicting, by the one or more processors, an outcome of the large-scale product test based at least on the first result and the second result. ### Claim 2 (depends on 1) The computer-implemented method of claim 1, wherein the small-scale product test is administered by a cone calorimeter. ### Claim 3 (depends on 1) The computer-implemented method of claim 1, further comprising: transmitting, by the one or more processors, the outcome, the first result, and the second result to a user device; and updating, by the one or more processors, the plurality of machine learning models using the outcome, the first result, and the second result. ### Claim 4 (depends on 1) The computer-implemented method of claim 1, further comprising: training, by the one or more processors, the plurality of machine learning models using an initial dataset indicating at least an initial set of results of (i) an initial set of large-scale products tested according to the large-scale product test, and (ii) an initial set of small-scale products tested according to the small-scale product test. ### Claim 5 A system for predicting an outcome of a large-scale product test, the system comprising: one or more processors; a memory storing data associated with a plurality of machine learning models; and a non-transitory computer-readable memory interfaced with the one or more processors, and storing instructions thereon that, when executed by the one or more processors, cause the one or more processors to: receive a set of small-scale results of a product tested according to a small-scale product test representative of the large-scale product test, calculate, based on the set of small-scale results as a first input to a first machine learning model of a plurality of machine learning models, a first result predicting an outcome of the product tested according to the large-scale product test, wherein calculating the first result includes: determining a first classification for the product, and calculating a confidence value for the first classification, calculate, based on the set of small-scale results as a second input to at least one second machine learning model of the plurality of machine learning models, a second result predicting the outcome of the product tested according to the large-scale product test, wherein calculating the second result includes: determining, based on a test profile for the large-scale product test, a second classification for the product, wherein determining the second classification includes: predicting, by the one or more processors and based on the set of small-scale results as the second input to the at least one second machine learning model of the plurality of machine learning models, the test profile, wherein: the at least one second machine learning model includes a plurality of regression models, and each regression model of the plurality of regression models predicts a test value of a plurality of test values of the test profile, the plurality of test values including at least one of: (i) a flame spread value; (ii) a total heat release value; (iii) a peak heat release rate value; and (iv) a fire growth rate index value, and predict an outcome of the large-scale product test based at least on the first classification, the confidence value, and the second classification. ### Claim 6 (depends on 5) The system of claim 5, wherein the small-scale product test is administered by a cone calorimeter. ### Claim 7 (depends on 5) The system of claim 5, wherein the non-transitory computer-readable memory stores additional instructions thereon that, when executed by the one or more processors, further cause the one or more processors to: transmit the outcome, the first result, and the second result to a user device; and train the plurality of machine learning models using the outcome, the first result, and the second result. ### Claim 8 (depends on 5) The system of claim 5, wherein the non-transitory computer-readable memory stores additional instructions thereon that, when executed by the one or more processors, further cause the one or more processors to: train the plurality of machine learning models using an initial dataset indicating at least an initial set of results of (i) an initial set of large-scale products tested according to the large-scale product test, and (ii) an initial set of small-scale products tested according to the small-scale product test. ### Claim 9 A non-transitory computer-readable memory storing one or more instructions for predicting an outcome of a large-scale product test that, when executed by one or more processors, cause the one or more processors to: receive a set of small-scale results of a product tested according to a small-scale product test representative of the large-scale product test; calculate, based on the set of small-scale results as a first input to a first machine learning model of a plurality of machine learning models, a first result predicting an outcome of the product tested according to the large-scale product test, wherein calculating the first result includes: determining a first classification for the product, and calculating a confidence value for the first classification; calculate, based on the set of small-scale results as a second input to at least one second machine learning model of the plurality of machine learning models, a second result predicting the outcome of the product tested according to the large-scale product test, wherein calculating the second result includes: determining, based on a test profile for the large-scale product test, a second classification for the product, wherein determining the second classification includes: predicting, by the one or more processors and based on the set of small-scale results as the second input to the at least one second machine learning model of the plurality of machine learning models, the test profile, wherein: the at least one second machine learning model includes a plurality of regression models, and each regression model of the plurality of regression models predicts a test value of a plurality of test values of the test profile, the plurality of test values including at least one of: (i) a flame spread value; (ii) a total heat release value; (iii) a peak heat release rate value; and (iv) a fire growth rate index value; and predict an outcome of the large-scale product test based at least on the first classification, the confidence value, and the second classification. ### Claim 10 (depends on 9) The non-transitory computer-readable memory of claim 9, wherein the small-scale product test is administered by a cone calorimeter. ### Claim 11 (depends on 9) The non-transitory computer-readable memory of claim 9, wherein the instructions further include additional instructions that, when executed by the one or more processors, cause the one or more processors to: transmit the outcome, the first result, and the second result to a user device; and train the plurality of machine learning models using the outcome, the first result, and the second result.