Background – What is Connected Operations on Cloud

Manufacturing enterprises have always focused on improving their operational technologies as times change, adopting automation in industrial control systems such as SCADA, Distributed Control Systems (DCS), Computer Numerical Controls (CNC) etc. Information Systems (IT) were always utilized by these enterprises separate from and complementary to OT, for communication and analytic purposes. But the evolution in Information Systems technologies, primarily the Internet of Things (IoT), Artificial Intelligence and Machine Learning, Cloud and Serverless Computing have necessitated the increased convergence of IT and OT technologies. The Industry 4.0, a conceptual framework, emphasizes this need of IT-OT convergence to build/transform manufacturing operations into a smart factory. Infosys, as a leader in IT services combining the Manufacturing domain expertise including various subdomains such as Automotive and Aerospace, has developed the Connected Operations on Cloud – a solution built on Industrial IoT platform. The solution enables industrial manufacturers harness IoT and cloud technologies to achieve smart manufacturing.

Challenges and Drivers in Automotive Industry

As we have seen before, enterprises belonging to various manufacturing domains to name a few – Automotive, Aerospace, etc. face the challenge to embrace latest IT developments and transform their existing operations into digital factory. Automotive industry has its unique set of challenges: any automotive enterprise spanning across geographical locations is divided into business function units known as shop with each having a different set of equipment, infrastructure and performance metric definitions. While some of these are common to other domains, the Automotive industry has a larger pressing need to embrace transformation since the demand for production of eco-friendly vehicles consuming alternate energy such as electric battery and Autonomous, connected vehicles increasing automation in vehicle control.

Any solution addressing the digital transformation need for Automotive industry has to cover both scenarios of development: Greenfield – a completely new setup involving IoT technologies and achieving faster IT-OT convergence and brownfield – a way of transforming operations while adapting to legacy systems and data. The Infosys Connected Operations on Cloud solution addresses these challenges but also has a template on specific domain’s challenges and in this case, has built an Automotive Manufacturing Template focused on solving Automotive customers’ needs.

The primary challenges of automotive enterprises in adopting smart manufacturing:

1. High Operating expenses: All the hardware infrastructure and software systems setup on customers’ premises would be unable to scale up or down based on demand and production lots. The number of different proprietary technologies and software products used to achieve the same business goals adds to rising operational costs.
2. High cost of Maintenance: If customers had maintenance of their assets in traditional reactive approach with no ability to predict based on past data, the scheduled and unplanned repairs of assets results in longer downtime of production operations and high costs.
3. Limited Visibility: The manufacturing IT and OT applications generally varies across business units and limited visibility for executives and business into plant floor operations.
4. Lack of organization-wide process harmonization: Each business unit generally sets up its own unique asset hierarchy, processes and KPI definitions. This means there are dependencies on unit-specific people to communicate with other stake holders for a coherent process and decision making at organization level. This leads to not seeing the challenges at a larger picture and making uninformed decisions.
5. Loss in traceability & compliance: The use of different applications running siloed in IT-OT systems means extra effort to meet compliance standards and maintain traceability often not met to expectations.

The key business drivers to overcome above challenges and enable smart factory for Automotive industry are:

1. Provide Scalability and Reduce Operational costs.
2. Improve Operational Performance with Predictability.
3. Improve Visibility and Sustainability
4. Provide Standardization and Customization
5. Provide Resilience and Reliability

Business Solution Approach for Automotive Smart Factory

Infosys follows a three-phased approach in solutioning a smart factory for any automotive customer:

1. Maturity Assessment of machines, sensors and communication standards in existing plants using Infosys Machine Categorization Framework into Cat 1 (Legacy Machines), Cat 2 (Analog Machines), Cat 3 (PLC and CNC Machines) and Cat 4 (Smart Machines).
2. Solution and Roadmap recommendations:  Infosys decide with business unit stakeholders on the initial plant POCs and establish overall roadmap for the customer.
3. Design and Pilot Implementation with Scaled rollout on a cloud platform leveraging Infosys Connected Operations on Cloud solution ensuring we meet Global Standards for Design, Architecture, Deployment and Processes. The outcome of pilot implementation gives a way to apprise and adjust the solution when implementing to other business units.

Infosys implements the business solution to transform operations of Automotive enterprises in the following ways:

• Build a system connecting the IT-OT world with End-to-End integration of Machines, People & Systems.
• Design a Hybrid Edge/Cloud Architecture to achieve reliable visibility and connectivity even when cloud network is not accessible at customer plant due to WAN disruptions. The Edge components are linked with cloud components using IoT services.
• Build a responsive and interactive persona-based Dashboard with customizable Hierarchy and KPI setup from plant to machine levels.
• Integration of Equipment data collection systems with other customer systems such as Maintenance, Quality, Energy Management to achieve a cohesive solution and reap better benefits.
• Adopt a Templatized approach such as using a standard framework like AWS Connected Device Framework for machine onboarding into the system and DevOps pipelines for CI/CD.

Key Components of Automotive Manufacturing Template

Infosys’ Connected Operations on Cloud solution can be implemented on any cloud platform and designed to be cloud provider-agnostic but with AWS we have our objectives aligned and well-integrated with AWS’ Industrial Data Fabric (IDF) to help customers define and understand the value of transforming manufacturing and industrial operations by applying a proven data driven approach to achieving business outcomes. In our 5 key components of the Automotive Manufacturing Template, we will see how we adopt the standard architectural principles and use best-of-breed AWS IoT services in any customer solution.

1. Reference Architecture: Hybrid Edge/Cloud Architecture and best of the AWS services with the advantages of Scalability on demand, high availability, reliability and stability. There are various AWS services to be leveraged in the solution such as AWS IoT Core and IoT Greengrass services for Industrial data collection, IoT SiteWise for industrial data modeling and metric computation, AWS Lambda functions, Amazon API Gateway, AWS Amplify and RDS Aurora database service for Serverless components, Amazon S3 for Storage, Amazon Athena for Analytics, Amazon SageMaker and Lookout for Equipment to build AI/ML prediction models. These components used in an efficient architecture enable the cloud advantage for the customer.

2. Hierarchy Model and KPI mapping: The application is designed to enable the customer define asset hierarchy and KPIs for each business unit allowing seamless onboarding of plant and equipment in bulk into the system. Shop-specific Hierarchy models and KPI definitions are setup for each business unit by setting up IoT SiteWise Asset Models. This allows organization-wide process harmonization while enabling customization across plants. SiteWise Metrics support KPI to be aggregated and computed per time windows avoiding writing boilerplate code. The inbuilt metric computation capabilities such as SiteWise Metrics may not always fit the need of customer metric definitions (such as requiring inputs from legacy systems not integrated into the solution) and may require writing complex code running as AWS Lambda functions.

3. IIoT Edge Data Collection: While the cloud platform and services provide resilience, the IoT edge setup provides reliable connectivity within customer network to real-time data and insights. Data collected from connected machines can be from different sources such as traditional databases such as Oracle and SQL Server and direct OPC-UA middleware integration with IoT SiteWise for information from PLC tags. The AWS IoT Greengrass components facilitate performing custom logic such as building database connectors to legacy databases such as Oracle, SQL Server etc. which transform records into data streams to be transported in AWS IoT SiteWise Gateways to Cloud for further processing. MQTT topics are used in IoT Core Rules engine to facilitate cheaper and reliable data transmission.

4. Predictive Maintenance: This key solution component is one of key differentiators in legacy vs smart manufacturing by empowering the customer to save operating expenses with accurately predicted equipment breakdowns. It is implemented using a host of AWS services such as Amazon SageMaker, Lookout for Equipment (L4E), Kinesis firehose, Lambda and S3 by:
a. Building machine learning (ML) models by ingesting historical sensor data in Amazon SageMaker and onboarding them into Amazon L4E service.
b. Near Real-Time Data processing and Transformation of raw equipment data from IoT SiteWise to L4E ML models to predict anomaly inferences.
c. Writing back prediction inferences from L4E to IoT SiteWise to display in the dashboard.
d. Ingesting maintenance work order data as feedback loop to train L4E ML model on periodic basis.

5. UI/UX Templates: Develop a dashboard responsive to render in mobiles, tabs and desktops. Infosys UX/UI templates are reused saving development time and themed according to customer needs. The dashboard contains templates for each hierarchy level and user persona with interactive controls such as Reports, Trends, Heatmap of assets needing attention.

Infosys Connected Operations on Cloud delivers these benefits to automotive customers by implementing Automotive manufacturing template:

 Conclusion

While the Automotive Manufacturing Template of Infosys Connected Operations solution addresses the major challenges of Automotive industry, the actual solution will have to consider each customer’s unique set of challenges and whether the solution is greenfield or brownfield. There is always an added challenge in a brownfield implementation such as developing adaptors to connect to legacy database and middleware holding contextual data but there is also an advantage that the legacy systems are tried and tested resulting in more reliable business decision making than a greenfield implementation. In any solution, the best approach is always to find a middle ground in choosing provisioned vs. serverless resources and choose an optimal decision on how much data should be available at edge (at customer premise) vs cloud thereby achieving the best of both worlds i.e., Reliability and Security at Edge with Scalability and Availability at Cloud.