# US 2026/0228635 A1 · Propagating authored packages to asset management platforms

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2026/0228635 A1
- Document ID: US20260228635A1
- Status: Published application
- Kind: A1
- Application number: 19/577,921
- Filed: 2026-03-25
- Published: 2026-08-06
- Earliest claimed: 2023-08-21
- Assignees: MaintainX Inc.
- Applicants: MaintainX Inc.
- Inventors: Stefan Cristian Turlica; Hugo Dozois-Caouette; Mathieu Marengère-Gosselin
- Web page: https://patentgrep.com/patent/US20260228635A1

## CPC

- G06F 8/65: Updates (security arrangements therefor G06F21/57)
- G06N 20/00: Machine learning
- G06V 30/10: Character recognition
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING

## Abstract

Bitemporal modeling is used to track availability of an asset. A machine learning (ML) model is instructed to train on the availability of the asset during a first training stage. After the first training stage and before a second training stage, the ML model is instructed to generate a prediction regarding future availability. After, data reflecting actual availability is collected and the prediction is compared against the data reflecting actual availability. The ML model is further trained to correct for any discrepancies between the prediction and the data reflecting actual availability. After the second training stage, an updated package is generated for an asset management platform that manages the asset.

## Claims

### Claim 1

A method, comprising:
using bitemporal modeling to track availability of an asset, wherein the bitemporal modeling includes:
a first record associated with the availability of the asset at a first point in time, from a perspective of the first point in time; and
a second record associated with the availability of the asset at the first point in time, from a perspective of a later point in time after the first point in time;
instructing a machine learning (ML) model to train on the availability of the asset that is tracked by the bitemporal modeling during a first training stage;
after the first training stage and before a second training stage, instructing the ML model to generate a prediction regarding a future availability of the asset, including at a second point in time that has yet to occur;
after the second point in time has occurred, collecting data reflecting an actual availability of the asset, including at the second point in time;
comparing the prediction regarding the future availability of the asset, including at the second point in time, against the data reflecting the actual availability of the asset, including at the second point in time;
instructing the ML model to further train on said comparison during a second training stage in order to correct for any discrepancies between the prediction regarding the future availability of the asset and the data reflecting the actual availability of the asset; and
after the second training stage, generating an updated package, wherein: at an asset management platform that manages the asset, the updated package is installed such that management of the asset by the asset management platform is modified, including one or more of the following: operation of the asset, maintenance of the asset, or use of the asset.

### Claim 2 (depends on 1)

The method of claim 1, wherein at least one of the first record or the second record is obtained from one or more of the following: a text source, a procedure, a camera, or a sensor.

### Claim 3 (depends on 1)

The method of claim 1, wherein:
the first training stage includes determining a standard operating range for the asset; and
generating the prediction includes determining whether at least one of the first record or the second record indicates anomalous behavior for the asset.

### Claim 4 (depends on 1)

The method of claim 1, further including: after the first training stage and before the second training stage, instructing the ML model to generate a recommended maintenance event for the asset.

### Claim 5 (depends on 4)

The method of claim 4, wherein the recommended maintenance event includes replacing a part of the asset.

### Claim 6

A system, comprising:
a processor that:
uses bitemporal modeling to track availability of an asset, wherein the bitemporal modeling includes:
a first record associated with the availability of the asset at a first point in time, from a perspective of the first point in time; and
a second record associated with the availability of the asset at the first point in time, from a perspective of a later point in time after the first point in time;
after a second point in time has occurred, collects data reflecting an actual availability of the asset, including at the second point in time;
compares a prediction regarding a future availability of the asset, including at the second point in time, against the data reflecting the actual availability of the asset, including at the second point in time; and
after a second training stage, generates an updated package, wherein: at an asset management platform that manages the asset, the updated package is installed such that management of the asset by the asset management platform is modified, including one or more of the following: operation of the asset, maintenance of the asset, or use of the asset; and
an interface to a machine learning (ML) model that:
instructs the ML model to train on the availability of the asset that is tracked by the bitemporal modeling during a first training stage;
after the first training stage and before the second training stage, instructs the ML model to generate the prediction regarding the future availability of the asset, including at the second point in time that has yet to occur; and
instructs the ML model to further train on said comparison during the second training stage in order to correct for any discrepancies between the prediction regarding the future availability of the asset and the data reflecting the actual availability of the asset.

### Claim 7 (depends on 6)

The system of claim 6, wherein at least one of the first record or the second record is obtained from one or more of the following: a text source, a procedure, a camera, or a sensor.

### Claim 8 (depends on 6)

The system of claim 6, wherein:
the first training stage includes determining a standard operating range for the asset; and
generating the prediction includes determining whether at least one of the first record or the second record indicates anomalous behavior for the asset.

### Claim 9 (depends on 6)

The system of claim 6, wherein the interface to the ML model further: after the first training stage and before the second training stage, instructs the ML model to generate a recommended maintenance event for the asset.

### Claim 10 (depends on 9)

The system of claim 9, wherein the recommended maintenance event includes replacing a part of the asset.
