Modeling your industrial property at scale utilizing AWS IoT SiteWise

Introduction

Industrial and manufacturing prospects more and more depend on AWS IoT SiteWise to gather, retailer, manage, and monitor information from industrial tools at scale. AWS IoT SiteWise supplies an industrial information basis for distant tools monitoring, efficiency monitoring, detecting irregular tools habits, and help for superior analytics use circumstances.

Constructing similar to a knowledge basis usually entails modeling your property and ingesting dwell and historic telemetry information. This may increasingly require a big effort when addressing tens of 1000’s of apparatus and ever-changing operations in pursuit of lowering waste and bettering effectivity.

We launched three new options for AWS IoT SiteWise at re:Invent 2023 to enhance your asset modeling efforts. Clients can now characterize tools elements utilizing Asset mannequin elements, selling reusability. With Metadata bulk operations, they will mannequin their tools and handle adjustments in bulk. Person-defined distinctive identifiers assist prospects obtain consistency throughout the group by utilizing their very own identifiers.

On this weblog submit, we are going to study 11 real-world buyer situations associated to asset modeling. We are going to share code examples that will help you be taught extra in regards to the new AWS IoT SiteWise options associated to every state of affairs.

Stipulations

1. Familiarity with asset modeling in AWS IoT SiteWise
2. An AWS account
3. Fundamental information of Python

Setup the surroundings

First, you’ll configure your developer workstation with AWS credentials and confirm that Python is put in. Subsequent you’ll set up Git, clone the code instance undertaking to your workstation, and arrange the undertaking. Lastly, you’ll create an AWS Id and Entry Administration (IAM) coverage.

1. Create a Cloud9 surroundings utilizing Amazon Linux 2 platform (really useful) or use any on-premises machine as a developer workstation
2. Configure AWS credentials
3. Confirm Python 3.x is put in in your system by working python3 --version or python --version (on Home windows)
4. Utilizing terminal, set up Git and clone the Metadata Bulk Operations Pattern for AWS IoT SiteWise repository from the AWS Samples library on Github

   sudo yum set up git
   git --version
   git clone https://github.com/aws-samples/metadata-bulk-operations-sample-for-aws-iot-sitewise.git

5. Set up required Python packages by working pip3 set up -r necessities.txt
6. Replace config/project_config.yml to offer required data for the job
   • s3_bucket_name: Identify of the S3 bucket the place bulk definitions will probably be saved
   • job_name_prefix: Prefix for use for the majority operations jobs
7. Create an AWS Id and Entry Administration (IAM) coverage with permissions that permit the alternate of AWS assets between Amazon S3, AWS IoT SiteWise, and your native machine. This can permit you to carry out bulk operations.

Onboard and handle property at scale

AWS IoT SiteWise now helps the majority import, export, and replace of business tools metadata for modeling at scale. These bulk operations are accessible via new API endpoints similar to CreateMetadataTransferJob, ListMetadataTransferJobs, GetMetadataTransferJob and CancelMetadataTransferJob.

With this new functionality, customers can bulk onboard and replace property and asset fashions in AWS IoT SiteWise. They’ll additionally migrate property and asset fashions between totally different AWS IoT SiteWise accounts.

You’ll primarily use metadata bulk import jobs for this weblog. The next diagram and steps clarify the workflow concerned in a metadata bulk import job.

Steps in Metadata Bulk Import Movement

1. Put together a job schema JSON file for AWS IoT SiteWise assets. This would come with asset fashions and property, following the AWS IoT SiteWise metadata switch job schema. Add this file to an Amazon S3 bucket.
2. Make a metadata bulk import name to AWS IoT SiteWise, referencing the uploaded JSON file
3. AWS IoT SiteWise will import all of the assets specified within the JSON file
4. Upon completion, AWS IoT SiteWise will return the standing and a presigned Amazon S3 URL for any failures encountered
5. If there are failures, entry the supplied report to research and perceive the basis trigger

You too can carry out bulk operations utilizing the console by navigating to Construct → Bulk Operations. Now that you simply perceive how metadata bulk operations work, you will note how this characteristic might help within the following real-world situations.

Situation 1 – Onboard preliminary asset fashions & property

Throughout a Proof of idea (POC), our prospects usually onboard a subset of their tools to AWS IoT SiteWise. Utilizing metadata bulk operations, you possibly can import 1000’s of asset fashions and property to AWS IoT SiteWise in a single import job.

For a fictitious automotive manufacturing firm, import asset fashions and property associated to the welding strains at one in every of its manufacturing crops.
python3 src/import/primary.py --bulk-definitions-file 1_onboard_models_assets.json

Situation 2 – Outline asset hierarchy

As soon as the asset fashions and property are created in AWS IoT SiteWise, you possibly can outline the connection between property and create an asset hierarchy. This hierarchy helps customers to trace efficiency throughout totally different ranges, from the tools degree to the company degree.

Create an asset hierarchy for Sample_AnyCompany Motor manufacturing firm
python3 src/import/primary.py --bulk-definitions-file 2_define_asset_hierarchy.json

Situation 3 – Affiliate information streams with asset properties

Our prospects usually begin ingesting information from their information sources such OPC UA server, even earlier than modeling their property. In these conditions, the information ingested into SiteWise is saved in information streams that aren’t related to any asset properties. As soon as the ingestion train is full, you need to affiliate the information streams with particular asset properties for contextualization.

Affiliate the information streams for Sample_Welding Robotic 1 and Sample_Welding Robotic 2 with corresponding asset properties.

python3 src/import/primary.py --bulk-definitions-file 3_associate_data_streams_with_assets.json

On this weblog, we created three separate metadata bulk import jobs. These jobs had been for creating asset fashions and property, defining the asset hierarchy, and associating information streams with asset properties. You too can carry out all of those actions utilizing a single metadata bulk import job.

Situation 4 – Onboard extra property

After demonstrating the enterprise worth throughout POC, the subsequent step is to scale the answer inside and throughout crops. This scale can embody remaining property in the identical plant, and new property from different crops.

On this state of affairs, you’ll onboard extra welding robots (#3 and #4), and a brand new manufacturing line (#2) from the identical Chicago plant.
python3 src/import/primary.py --bulk-definitions-file 4_onboard_additional_assets.json

Situation 5 – Create new properties

You possibly can improve asset fashions to accommodate adjustments in information acquisition. For instance, when new sensors are put in to seize extra information, you possibly can replace the corresponding asset fashions to replicate these adjustments.

Add a brand new property Joint 1 Temperature to Sample_Welding Robotic asset mannequin
python3 src/import/primary.py --bulk-definitions-file 5_onboard_new_properties.json

Situation 6 – Repair handbook errors

Errors can happen throughout asset modeling particularly when customers manually enter data. Examples embody asset serial numbers, asset descriptions, and models of measurement. To right these errors, you possibly can replace the data with the right particulars.

Right the serial variety of Sample_Welding Robotic 1 asset by changing the outdated serial quantity S1000 with S1001.
python3 src/import/primary.py --bulk-definitions-file 6_fix_incorrect_datastreams.json

Situation 7 – Relocate property

Manufacturing line operations change for a number of causes, similar to course of optimization, technological developments, and tools upkeep. Because of this, some tools might transfer from one manufacturing line to a different. Utilizing Metadata bulk operations, you possibly can replace the asset hierarchy to adapt to the adjustments in line operations.

Transfer Sample_Welding Robotic 3 asset from Sample_Welding Line 1 to Sample_Welding Line 2.
python3 src/import/primary.py --bulk-definitions-file 7_relocate_assets.json

Situation 8 – Backup asset fashions and property

AWS recommends that you simply take common backups of asset fashions and property. These backups can be utilized for catastrophe restoration or to roll again to a previous model. To create a backup, you should utilize the bulk export operation. Whereas exporting, you possibly can filter particular asset fashions and property to incorporate in your exported JSON file.

You’ll now again up the definitions of all welding robots beneath welding line 1. Substitute <YOUR_ASSET_ID> in 6_backup_models_assets.json with the Asset ID of Sample_Welding Line 1.

python3 src/export/primary.py --job-config-file 8_backup_models_assets.json

Situation 9 – Promote asset fashions and property to a different surroundings

Through the use of the metadata bulk export operation adopted by the majority import operation, you possibly can promote a set of asset fashions and property from one surroundings to a different.

Promote all of the asset fashions and property from the event to the testing surroundings.
python3 src/import/primary.py --bulk-definitions-file 9_promote_to_another_environment.json

Preserve consistency all through the group

Many industrial corporations might have modeled some or most of their industrial tools in a number of techniques similar to asset administration techniques and information historians. It’s important for these corporations to make use of widespread identifiers throughout the group to keep up consistency.

AWS IoT SiteWise now helps using exterior ID and user-defined UUID for property and asset fashions. With the exterior ID characteristic, customers can map their present identifiers with AWS IoT SiteWise UUIDs. You possibly can work together with asset fashions and property utilizing these exterior IDs. The user-defined UUID characteristic helps customers to reuse the identical UUID throughout totally different environments similar to improvement, testing, and manufacturing.

To be taught in regards to the variations between exterior IDs and UUIDs, consult with exterior IDs.

Situation 10 – Apply exterior identifiers

You possibly can apply exterior IDs utilizing the AWS IoT SiteWise console, API, or metadata bulk import job. This may be executed for present asset fashions, or property with none exterior IDs in AWS IoT SiteWise.

Apply exterior ID to an present asset, for instance, Sample_Welding Robotic 4.
python3 src/import/primary.py --bulk-definitions-file 10_apply_external_identifier.json

Promote standardization and reusability utilizing mannequin composition

AWS IoT SiteWise launched help for a part mannequin. That is an asset mannequin kind that helps industrial corporations mannequin smaller items of apparatus and reuse them throughout asset fashions. This helps standardize and reuse widespread tools elements, similar to motors.

For instance, a CNC Lathe (asset mannequin) is fabricated from elements similar to servo motors. With this characteristic, a servo motor may be modeled independently as a part and reused in one other asset mannequin, similar to a CNC Machining Middle.

Situation 11 – Compose asset fashions

You possibly can compose asset fashions utilizing the AWS IoT SiteWise console, API or metadata bulk import job.

Compose the Sample_Welding Robotic asset mannequin by independently modeling elements in a welding robotic, similar to a robotic joint.
python3 src/import/primary.py --bulk-definitions-file 11_compose_models.json

Clear Up

When you now not require the pattern resolution, think about eradicating the assets.

Run the next to take away all of the asset fashions and property created utilizing this pattern repository.
python3 src/remove_sitewise_resources.py --asset-external-id External_Id_Company_AnyCompany

Conclusion

On this submit, we demonstrated using new AWS IoT SiteWise options, similar to Metadata bulk operations, Person-defined distinctive identifiers, and Asset mannequin elements. Collectively, these options promote standardization, reusability, and consistency throughout your group, whereas serving to you to scale and improve your asset modeling initiatives.

Concerning the authors

Raju Gottumukkala is a Senior WorldWide IIoT Specialist Options Architect at AWS, serving to industrial producers of their sensible manufacturing journey. Raju has helped main enterprises throughout the power, life sciences, and automotive industries enhance operational effectivity and income progress by unlocking true potential of IoT information. Previous to AWS, he labored for Siemens and co-founded dDriven, an Business 4.0 Knowledge Platform firm.