US-12626205-B2 - Method and system for semi-automatic completion of an engineering project

US12626205B2US 12626205 B2US12626205 B2US 12626205B2US-12626205-B2

Abstract

An initial sequence representing a partially configured engineering project is processed by a recurrent neural network to generate recommendations being a sequence of complementary items that completes an engineering project. A feature predictor component computes a set of features for each recommendation. A bisection component selects a feature from the sets of features that distinguishes some of the recommendations and forms pruned recommendations by choosing all instances from the recommendations that have the selected feature. A user interface displays the selected feature, detects a user interaction indicating that the selected feature is required, outputs the pruned recommendations. The engineering project is completed by combining the initial sequence with the chosen pruned recommendation. As a result, a user is supported in choosing optimal modules, as the selected feature can distinguish the recommendations that have the desired technical properties or target system KPI.

Inventors

Serghei Mogoreanu
Marcel Hildebrandt

Assignees

SIEMENS AKTIENGESELLSCHAFT

Dates

Publication Date: 20260512
Application Date: 20220615
Priority Date: 20210628

Claims (10)

1 . A method for semi-automatic completion of an engineering project, comprising the following operations performed by components, wherein the components are software components executed by one or more processors and/or hardware components: calculating, by a first artificial intelligence component of a recommendation engine comprising one or more processors, latent representations of initial items of an initial sequence, with the initial sequence representing a partially configured engineering project, with the initial items representing hardware modules and/or software modules used in the engineering project, wherein the first artificial intelligence component comprises a trained feature learning component adapted to calculate the latent representations of the initial items, and is trained on item properties stored in a first database storing the item properties of a plurality of different items, the first database accessible by the recommendation engine; encoding, by the first artificial intelligence component, technical information of the hardware and/or software modules into the latent representations; processing, by a second artificial intelligence component of the recommendation engine, the latent representations to generate recommendations, with each recommendation being a sequence of complementary items that completes the engineering project when combined with the initial sequence, wherein the second artificial intelligence component comprises a trained sequential model adapted to calculate the sequence of complementary items as the recommendations, and is trained on historically completed item sequences stored in a second database storing the historical completed item sequences, the second database accessible by the recommendation engine; computing, by a feature predictor component, a set of features for each recommendation, wherein the feature predictor component predicts technical features of each recommendation generated by the second artificial intelligence component; selecting, by a bisection component, a feature from the sets of features that distinguishes some of the recommendations; displaying, by a user interface, the selected feature, and detecting, by the user interface, a user interaction indicating that the selected feature is required; forming pruned recommendations by choosing all instances from the recommendations that have the selected feature; outputting, by the user interface, the pruned recommendations, and detecting, by the user interface, a user interaction choosing one of the pruned recommendations; automatically completing the engineering project by combining the initial sequence with the chosen pruned recommendation; in response to the automatically completing the engineering project, updating the second database with the completed engineering project such that the completed engineering project is used as additional training of the second artificial intelligence component so that the second artificial intelligence component improves over time as a function of the updating; outputting and/or storing the completed engineering project; and automatically producing the engineering project according to the completed project.
2 . The method according to claim 1 , wherein: the second artificial intelligence component has been trained to assign a plausibility score to each recommendation, indicating the plausibility of the respective sequence of complementary items, and the pruned recommendations are sorted in the outputting operation according to their plausibility score on a display of the user interface.
3 . The method according to claim 1 , wherein: the feature predictor component includes a rule-based system, and/or the feature predictor component includes at least one simulation module that computes at least some of the features, and/or at least some of the features are key performance indicators including energy efficiency measures and/or performance measures.
4 . The method according to claim 1 , wherein the bisection component is adapted to select a feature that distinguishes approximately half of the recommendations by choosing a feature that maximizes a Gini index.
5 . The method according to claim 1 , wherein the second artificial intelligence component is implemented as a sequence-to-sequence model, and wherein the sequence-to-sequence model processes the input sequence.
6 . The method according to claim 1 , wherein the latent representations encode technical information about the hardware modules and/or software modules.
7 . The method according to claim 1 , wherein the pruned recommendations are distinguished from the recommendations in a corresponding visualization that shows at least one way the recommendations are conflicting.
8 . The method according to claim 1 , wherein each latent representation comprises a vector for different technical properties of the associated module, wherein the technical properties include supply voltage, fail-safe compatibility, power consumption, and screen resolution.
9 . A system for semi-automatic completion of an engineering project, comprising the following components, wherein the components are software components executed by one or more processors and/or hardware components: a first artificial intelligence component of a recommendation engine comprising one or more processors, adapted for calculating latent representations of initial items of an initial sequence, with the initial sequence representing a partially configured engineering project, and with the initial items representing hardware modules and/or software modules used in the engineering project, wherein the first artificial intelligence component comprises a trained feature learning component adapted to calculate the latent representations of the initial items, and is trained on item properties stored in a first database storing the item properties of a plurality of different items, the first database accessible by the recommendation engine, wherein the first artificial intelligence component encodes technical information of the hardware and/or software modules into the latent representations; a second artificial intelligence component of the recommendation engine, adapted for processing the latent representations to generate recommendations, with each recommendation being a sequence of complementary items that completes the engineering project when combined with the initial sequence, wherein the second artificial intelligence component comprises a trained sequential model adapted to calculate the sequence of complementary items as the recommendations, and is trained on historically completed item sequences stored in a second database storing the historical completed item sequences, the second database accessible by the recommendation engine; a feature predictor component, adapted for computing a set of features for each recommendation, wherein the feature predictor component predicts technical features of each recommendation generated by the second artificial intelligence component; a bisection component, adapted for selecting a feature from the sets of features that distinguishes some of the recommendations and for forming pruned recommendations by choosing all instances from the recommendations that have the selected feature; a user interface, adapted for: displaying the selected feature and detecting a user interaction indicating that the selected feature is required, outputting the pruned recommendations and detecting a user interaction choosing one of the pruned recommendations; and at least one hardware and/or software component, adapted for automatically completing the engineering project by combining the initial sequence with the chosen pruned recommendation, outputting and/or storing the completed engineering project, and, in response to the automatically completing the engineering project, updating the second database with the completed engineering project such that the completed engineering project is used as additional training of the second artificial intelligence component so that the second artificial intelligence component improves over time as a function of the updating.
10 . A computer program product, comprising a computer readable hardware storage device having computer readable program code stored therein, said program code executable by a processor of a computer system to implement a method comprising: calculating, by a first artificial intelligence component of a recommendation engine comprising one or more processors, latent representations of initial items of an initial sequence, with the initial sequence representing a partially configured engineering project, and with the initial items representing hardware modules and/or software modules used in a engineering project; encoding, by the first artificial intelligence component, technical information of the hardware and/or software modules into the latent representations; processing, by a second artificial intelligence component of the recommendation engine, the latent representations to generate recommendations, with each recommendation being a sequence of complementary items that completes the engineering project when combined with the initial sequence, wherein the first artificial intelligence component comprises a trained feature learning component adapted to calculate the latent representations of the initial items, and is trained on item properties stored in a first database storing the item properties of a plurality of different items; computing, by a feature predictor component, a set of features for each recommendation, wherein the feature predictor component predicts technical features of each recommendation generated by the second artificial intelligence component, wherein the second artificial intelligence component comprises a trained sequential model adapted to calculate the sequence of complementary items as the recommendations, and is trained on historically completed item sequences stored in a second database storing the historical completed item sequences; selecting, by a bisection component, a feature from the sets of features that distinguishes some of the recommendations; displaying, by a user interface, the selected feature, and detecting, by the user interface, a user interaction indicating that the selected feature is required; forming pruned recommendations by choosing all instances from the recommendations that have the selected feature; outputting, by the user interface, the pruned recommendations, and detecting, by the user interface, a user interaction choosing one of the pruned recommendations; automatically completing the engineering project by combining the initial sequence with the chosen pruned recommendation; in response to the automatically completing the engineering project, updating the second database with the completed engineering project such that the completed engineering project is used as additional training of the second artificial intelligence component so that the second artificial intelligence component improves over time as a function of the updating; and outputting and/or storing the completed engineering project.

Description

CROSS-REFERENCE TO RELATED APPLICATIONS This application claims priority to EP Application No. 21182127.7, having a filing date of Jun. 28, 2021, the entire contents of which are hereby incorporated by reference. FIELD OF TECHNOLOGY The following relates to a method and system for semi-automatically completing a complex engineering project, in particular an automation system. BACKGROUND Over the past two decades, advancements in industrial automation have transformed the factory floors and processes in a number of industries, ranging from process industries (e.g., Oil and Gas, Power and Utilities, Water and Wastewater) to hybrid industries (e.g., Food and Beverage, Wind) to discrete industries (e.g., Automotive Manufacturing, Aerospace, Robotics). Automating processes in any of these industries requires engineers to design and configure an industrial automation solution—in other words to complete an engineering project for a complex system, consisting of a multitude of individual modules, the interplay of which fulfills the functional requirements arising from the intended application. Selection of necessary modules is typically done using configuration software offered by their manufacturer. Other examples of engineering projects include printed circuit boards and autonomous vehicles. These systems, like the industrial automation solutions described above, can be complex and comprise a multitude of different modules. The configuration of the complex engineering projects may comprise an iterative process, in which a user incrementally selects modules (hardware and software components for building the engineering project). The combination of these selected modules can fulfill functional requirements of the engineering projects while being also compatible with one another. The configuration of a complex engineering process is not an easy task and requires time, effort, experience, and a certain amount of domain-specific knowledge to be completed correctly by a user. WO 2021037603 A1 discloses a recommendation engine to automatically provide recommendations in order to support a user in the completion of an engineering project. The entire contents of that document are incorporated herein by reference. SUMMARY An aspect relates to provide a method and system that provide an alternative to the state of the art. According to the method for semi-automatic completion of an engineering project, the following operations are performed by components, wherein the components are software components executed by one or more processors and/or hardware components: calculating, by a first artificial intelligence component, latent representations of initial items of an initial sequence, with the initial sequence representing a partially configured engineering project, and with the initial items representing hardware modules and/or software modules used in the engineering project, andprocessing, by a second artificial intelligence component, the latent representations to generate recommendations, with each recommendation being a sequence of complementary items that completes the engineering project when combined with the initial sequence. The method is characterized by the following operations that are performed by components, wherein the components are software components executed by one or more processors and/or hardware components: computing, by a feature predictor component, a set of features for each recommendation,selecting, by a bisection component, a feature from the sets of features that distinguishes some of the recommendations,displaying, by a user interface, the selected feature, and detecting, by the user interface, a user interaction indicating that the selected feature is required,forming pruned recommendations by choosing all instances from the recommendations that have the selected feature,outputting, by the user interface, the pruned recommendations, and detecting, by the user interface, a user interaction choosing one of the pruned recommendations,automatically completing the engineering project by combining the initial sequence with the chosen pruned recommendation, andoutputting and/or storing the completed engineering project. The system for semi-automatic completion of an engineering project comprises the following components, wherein the components are software components executed by one or more processors and/or hardware components: a first artificial intelligence component, adapted for calculating latent representations of initial items of an initial sequence, with the initial sequence representing a partially configured engineering project, and with the initial items representing hardware modules and/or software modules used in the engineering project, anda second artificial intelligence component, adapted for processing the latent representations to generate recommendations, with each recommendation being a sequence of complementary items that completes the engineering project when combined with the