Executive Summary

This use case addresses data acquisition, which represents a huge challenge in the industry. Working with PLC manufacturers is costly and time-consuming. By finding clever solutions for data acquisition, weeve can provide our customers with more options.

Purpose and Scope

Connecting a manufacturing shop floor to business systems is a lengthy process that leaves shop floor managers dependent on long IT work cycles. In order to visualize your machine data, you first need to be able to extract the data from your machines. Extracting the data from the machines means that your IT department will have to block out time to connect your machines to their IT systems. In many cases, these projects are postponed because the IT departments are too overloaded and because of the complexity of connecting to the machines' PLCs and querying the collected data. Is there any way to get around the dependency on IT? Can we avoid the connection to the PLCs of the machines to get interesting data about the performance of the machine in a simple and fast way?

Business Value

Your machines communicate with each other using different communication protocols. Intercepting these communications and translating them so that the data is actionable is a complex task. Even for the simplest insights, such as figuring out whether a machine is working or not, you need to translate these protocols if you are relying (depending) on the data provided by the machine. This means that the simplest insight from your machine may require a year-long project.

With the weeve solution, you can bypass the PLC manufacturer, connect your machines to extract data from your industrial machines, and gain insights from the machine data. Weeve overcomes the obstacles to data collection by combining industry-standard hardware, cutting-edge sensor technology, and world-class modular software. Weeve works with a variety of hardware partners whose technology is as commonplace as a mobile phone or as advanced as industry-certified specialty components. The hardware selected by our partners extracts production data from your machines in as little as 3 minutes. This is done with easy to install hardware with the latest sensor technologies available. This is made possible by commodity price pressure driving down key sensor technologies namely in the mobile sector. Retrofitting your machine has never been easier. You can choose to use an existing packaged solution that is ready to go. This comes with pre-installed software and a SIM card. The set-up is simple, so you can get started quickly. You can choose to use the solution provider's cloud to see special dashboard visualizations for the price of a monthly subscription. Furthermore, you can track your machine's production capabilities by tracking on/off cycles - without waiting for a 2-year integration project with your IT department.

Alternatively, you can customize your setup. For example, if you have "on-premise" networks set up where sending SIM data out of the factory floor is not possible. Another example is if you need to connect your machines to another subsystem, which implies having another challenging integration with your IT department. In either case, weeve offers alternatives.

Weeve's software is capable of mapping all normal IoT data from your PLCs into normal data streams that humans can understand. You can aggregate IoT data from your PLCs and enrich it with data from external sources to ensure the highest quality data, such as data with higher frequency or precision. Leave it to weeve to take care of the intricacies - you simply decide where you want the data to go to extract the most value possible.

Weeve gives you access to a data marketplace where you can choose from a library of existing solutions. Alternatively, you can create your own custom connectors to get your IoT data to the right business system. In addition, the weeve manager lets you deploy IoT applications on your edge infrastructure. This means: build once. Configure, deploy, and automate. Realize the possibilities you could only imagine before.

Solution Overview

Assumptions:

  • Internet connection in the PLC area (i.e. WiFi or Ethernet access) if partner’s hardware without SIM
  • No internet connection in the PLC area if partner’s hardware with SIM
  • If data is passed to another “on-premise” system, then the system should allow connection
  • PLC ID or another identification number

Architecture and Setup:

  • PLC needs to be connected to weeve partner’s hardware, so that it could read data (it can be either contactless “observing” PLC with cameras or vibration sensors, or it can be mounted directly to PLC)
  • PLC has specification required by weeve partner’s hardware
  • Weeve Node Service runs directly on weeve partner’s hardware. If such arrangement is not possible, then weeve Node Service runs on a separate hardware, i.e. Raspberry Pi, that is connected to weeve partner’s hardware
  • If PLC has already attached hardware to read its data, then weeve Node Service can be connected directly to that hardware
  • The hardware with weeve Node Service can access the internet so that it can send data to the cloud or generate alerts when needed.

Suggested Data Services:

The following data services are suggested solutions that could be easily modified to suit the needs of our customers. For instance, different data processing modules could be applied or data could be egressed to other external databases and repositories. Moreover, data could be output to other “on-site” systems or another notification medium could be implemented.


Case 1 → Filter data and write to an external database.

Weeve partner’s hardware reads data from PLC machine and passes to weeve Node Service. Our data service filters out data according to our customers’ requirements and writes that data to a selected database, i.e. InfluxDB or AWS DynamoDB.




Case 2 → Sanitize data and create alert

Data from weeve partner’s hardware are sanitized and anomalies are detected. Later alert notifications can be triggered by setting an appropriate threshold. Notifications can be sent to a chosen medium: Slack, Microsoft Teams, WeChat, SMS, WhatsApp, or Messenger.


Case 3 → Use of machine learning to predict danger and alert.

Data from PLC could be passed through a special customer customized weeve module that uses machine learning to predict future events and trigger notification alert. Such machine learning modules could be created in cooperation with other weeve partners.



Requirements:

Technical and Hardware

  • Constant connection to the internet for data transmission to alert mediums or databases
  • Hardware that reads data is connected to PLC (either provided by customers themselves or weeve partners)

Functional Requirements

  • Hardware connected to PLC or a separate hardware running weeve Node Service runs Docker
  • Runs weeve Node Service

Non-Functional Requirements

  • The system should be always running
  • The system should be able to update its data service when required


Constraints:

  • The above-mentioned machine learning model is not evolved to predict everything. This only demonstrates the possibility of running Machine Learning algorithms on the edge using the weeve ecosystem.
  • Privacy and Intellectual Property(IP): The amount of data that is being sent will be limited to data that can be generated and read by hardware connected to PLC (either provided by customers themselves or weeve partners)