Applied AI Engineer Designer at Nokia

Job Description

ENVIROMENT

MW Products are fully designed inside the business unit, and this means that both the hardware and the software are designed and developed inside the R&D.

HW and SW domains have different processes and the introduction of the AI/ML in these processes will increase the quality of the output produced and will reduce manual job.

The work will be distributed on two scenarios described below. **

WHAT NEDED And WHAT LEARNED**

The student needs to have basic competences on office 365 tools p (Excel, Access) and basic knowledge on software development: we will use Visual Studio and C# to build data model and configure the AI entities.

The student will learn how to use the Azure DevOps Services (for code review and debugging), the Azure AI Machine (for ML and related logics) and the usage of MS Automate/Workflow will be used to drive AI entities.

The student will see how a SW R&D is producing and testing the builds in a multisite environment (teams are in different towns).

Student will understand how the hardware team is defining Engineering/Manufacturing strictures and basic concepts of approval workflows associated to object managed in this area.

SCENARIO 1: AI APPLIED TO SW DEVELOPMENT & AUTOMATION

The picture below is showing current and future development and test process in SW-FW area.

In current process developers (A) produce code and store it into GIT (B) and they manually write the test scripts (C) which will run in the automation machine (D).

AT the end of automatic tests, the metadata (results) are saved in the MW Dataverse (F) and logs in the central storage or Data Warehouse (G).

The scope of this scenario is to setup and define instructions to activate/apply two AI machines:

• Entity 1 will read and parse the logs to create a metadata stream with the full set of output from the test which has generated the log;
• Entity 2 will read the code generated in GIT and it will produce the test scripts; this entity will have also a module of Machine learning as it will read the output created by the Entity 1 to re-adapt scripts which have not run correctly.

The output of the two entities will be used eventually by the AI engine of GIT Co-Pilot to support developers in our code creation.

The stage will be focused in creating the data model inside the MW Dataverse and to define the process followed by the two entities to perform the tasks and the activities described (Parsing, Coding and Learning).

SCENARIO 2: AI APPLIED TO COMPONENT CLASSIFICATION

Each product designed in the hardware team has a Bill of Material (BoM), an engineering structure (A) composed by part designed internally (P) and component (C).

Each component (B) has a classification (E) composed by nodes and attributes (physical, logical, economical).

Each component can have one of more manufacturers attributing to the component a Manufacturing code (C)

The current process is a manual process to read the characteristics of the component supplied via one or more datasheets supplied.

The result is that very few components have the classification node completed.

The scope of this scenario is to setup and teach an AI entity (1) how to load and parse from the MW PDM the datasheets supplied by each manufacturer.

The entity will then update the classification attributes for the part selected.

The stage will focus on the definition of the data model inside the MW Dataverse and to define the process followed by the entity to perform the parse and the compilation of classification nodes.

Generation of reports should be included as results of a correct and complete classification of components (electrical, mechanical). **

Responsibilities**

SCENARIO 1: AI APPLIED TO SW DEVELOPMENT & AUTOMATION

The stage will be focused in creating the data model inside the MW Dataverse and to define the process followed by the two entities to perform the tasks and the activities described (Parsing, Coding and Learning).

SCENARIO 2: AI APPLIED TO COMPONENT CLASSIFICATION

The stage will focus on the definition of the data model inside the MW Dataverse and to define the process followed by the entity to perform the parse and the compilation of classification nodes.

Generation of reports should be included as results of a correct and complete classification of components (electrical, mechanical). **

Required Skills**

Knowledge requested: MS Office, basic knowledge of C# (or any other development language).