Salesforce’s closer collaboration with AWS: what does it mean for IoT?
During the Trailheadx developer event on June 23, 2021, Salesforce announced it is expanding its partnership with Amazon Web Services (AWS). Salesforce will embed AWS services for voice, video, AI, and machine learning directly into new apps for sales, service, and industry vertical use cases, like IoT. An exciting evolution.
At Waylay, we firmly believe that this is another sign of a maturing convergence of IT and OT. In the context of IoT, AWS provides multiple services. In this blog post, we explain how Waylay Digital Twin closes the gap between an AWS based IoT infrastructure and a Salesforce based CRM environment and how Salesforce unlocks new business value by adding customer context to IoT data. In particular, Waylay Digital Twin already leverages the strengths of AWS Elastic infrastructure, AWS IoT Core, and AWS Sagemaker for effortless predictive and preventive care in Salesforce Service Cloud.
The Waylay Digital Twin App solution brings the following benefits on top of the basic cloud building blocks and services that AWS provides:
- Adheres to the no-code philosophy of Salesforce, avoiding low-level programming by IT in AWS cloud (e.g. lambda functions, step functions, etc). This leads to more agility and faster response to changing customer or market requirements.
- Gives the power of IoT data in the hands of business users for both IoT data visibility and rule creation for condition and predictive/preventive based maintenance strategies. This dramatically shortens the time-to-value and avoids an IT bottleneck.
- Provides the most advanced & flexible no-code rule configuration options
- Allows business users to leverage results of in-house AI/ML initiatives.
- Has the richest integration capability with IoT platforms, next to AWS IoT Core
Waylay Digital Twin, an ISV app available on the Salesforce AppExchange, relies on an off-core component called the Waylay Engine. This engine is a cloud function orchestration engine deployable in the Amazon Elastic Elastic Kubernetes Service (EKS) - which uses Cassandra (e.g., Amazon Keyspaces) as hot storage for the IoT data. The Waylay Engine is a unique and very scalable IoT data processing engine that is programmable via:
- A low-code rule designer console focussed on domain and data experts
- A no-code rule designer app focussed on the Salesforce business user
The Waylay Engine can seamlessly ingest IoT data from AWS IoT Core. For example, suppose you have IoT devices managed through AWS IoT Core. In that case, the Salesforce Admin can use the Waylay Digital Twin App to self-configure the Waylay Engine subscription to AWS Thing Shadow topics.
The Waylay Engine can automatically detect AWS Things (IoT devices) and automatically create a corresponding Waylay Digital Twin Asset and parameters in Salesforce, allowing the Salesforce business user to:
- Explore IoT data - real-time and historical - from within the business context
- Create IoT data processing rules that generate Waylay Alarms in Salesforce
Whenever a new IoT device comes online.
The Waylay Engine can also host pre-trained machine learning algorithms (Python, Tensorflow, …) or use AWS Sagemaker hosted AI/ML algorithms that process AWS IoT data. A data science team can build these algorithms, and the business user can use the output in the Salesforce context using the Waylay Digital Twin App.
Typical use cases
- Parameter value predictions based on history
- E.g. (Building Management) forecast the future energy consumption and energy cost of a building.
- Outlier alarming
- E.g. (Manufacturing) unsupervised anomaly detection of temperature outliers indicates an unbalanced production process.
- Pattern detection to generate low-level alarms
- E.g. (Utilities) Predict & identify future component failure by classification of sensor readings.
- Object recognition in images
- E.g. (Retail) Smart cameras detecting shelves with out-of-stock goods.
If the AI/ML algorithm is also hosted as a microservice on AWS Sagemaker, this microservice can post its predictions on an AWS SNS or AWS Kinesis pub/sub topic. Waylay Digital Twin allows the Admin to configure the Waylay Engine to listen to these prediction messages. Modeling these AI/ML algorithms as if they were ‘virtual sensor’ child assets of the physical asset can also be done with Waylay Digital Twin.
Modeling these AI/ML algorithms brings the following benefits to the business user:
- The business user can view the current and past AI/ML predictions from within the business context of the asset.
- The business user can create alarming rules when, e.g., predictions deviate too much from the actual measurements.
In short, Waylay Digital Twin supports a scalable and sustainable collaboration model between the business user and the data scientist. The latter can focus on creating the best AI/ML algorithms, while the former can use the output of these algorithms to apply them in the support and maintenance context of the customer’s contract. And the best of it: the business user does not need any IT development effort to do this.
Read more about Intelligent IoT
The Waylay Digital Twin App crosses the divide between the AWS IoT services and the Salesforce Service Cloud and Field Service eco-system. It enables the Salesforce business users to take full advantage of IoT data in the customer’s service context without worrying about the scalability and maintainability of IoT data processing logic.