IoT Tips

Those who have been working with ThingWorx for many years will have noticed the work done around ingress stress testing and performance optimization.  Adding InfluxDB as a time-series data persistence provider really helped level up these capabilities while simultaneously decreasing the overall resources required by the infrastructure.  However with this ease comes a hidden challenge: query and data processing performance to work it into something useful.   Often It's Too Much Data In general most customers that I work with want to collect far too much data -- without knowing what it will be used for, or what processing will be required in order to make it usable and useful.  This is a trap in general with how many people envision IoT projects, being told by infrastructure providers that cloud storage and compute resources are abundant and cheap and that they should get as much data as possible.  This buildup of data means that more effort needs to be spent working it into something useful (data engineering/feature extraction) and addressing common data issues (quality, gaps, precision, etc.).  This might be fine for mature companies with large data analytics teams; however this is a makeup that I've only seen in the largest of our customers.  Some advice - figure out what you need and how you'll use it, and then collect that.  Work on extracting value today rather than hoping that extra data collected  now will provide some insights years from now.   Example - Problem Statement You got your Thing Model designed, and edge devices connected.  Now you've got data flowing in and being stored every 5 seconds in InfluxDB.  Great progress!  Now on to building the applications which cover the various use cases. The raw data is most likely going to need to be processed and potentially even significantly transformed into other information in order to make it useful.  Turning a "powered on and running" BOOLEAN to an "hour meter" INTEGER is a simple example.  Then you may need to provide a report showing equipment run time hours by day over a month.  The maintenance team may also have asked to look for usage patterns which lead to breakdowns, requiring extracting other data points from the initial one like number of daily starts, average daily run time, average time between restarts. The problem here is that unless you have prepared these new data points and stored them as well (say in a Stream), you are going to have to build these data sets on the fly, and that can be time and resource intensive and not give you the response time expected.  As you can imagine, repeatedly querying and processing large volumes of unchanging raw data is going to have resource and time implications - so this is why data collection and data use need to be thought about separately.   Data Engineering In the above examples, the key is actually creating new data points which are calculated progressively throughout normal operation.  This not only makes the information that you want available when you need it - in the right format - but it also significantly reduces resource requirements by constantly reprocessing raw data.  It also helps managing data purging, because as you create and store usable insights, you can eventually just archive away your old raw data streams.   Direct Database Queries vs. Thingworx Data Services Despite the above being a rule of thumb, sometimes a simple well structured database query can get you exactly what you need and do so quite quickly.  This is especially true for InfluxDB when working with extremely large time-series datasets.  The challenge here is that ThingWorx persistence providers abstract away the complexity of writing ones own database queries, so we can't easily get at the databases raw power and are forced to query back more data than needed and work it into a usable format in memory (which is not fast).   Leveraging the InfluxDB API using the ContentLoader Technique As InfluxDBs API is 100% REST, we can access it using in-built ThingWorx Content Loader services.  Check out this demonstration and explanation video where I talk about how to interact directly with InfluxDB in order to crush massive time-series data and get back much more usable and manageable data sets.  It is important to note here that you should use a read-only database user here, as you should never modify the ThingWorx databases to avoid untested scenarios which may lead to data corruption.   Optimizing ThingWorx query performance with the InfluxDB REST API - YouTube InfluxToolBox ThingWorx demo project (by T. Wobben)      

I've had a lot of questions over the years working with Azure IoT, Kepware, and ThingWorx that I really struggled getting answers to. I was always grateful when someone took the time to help me understand, and now it is time to repay the favour.   People ask me many things about Azure (in a ThingWorx context), and one of the common ones has been about MQTT communications from Kepware to ThingWorx using IoT Hub. Recently the topic has come up again as more and more of the ThingWorx expert community start to work with Azure IoT. Today, I took the time to build, test, validate, and share an approach and utilities to do this in cases where the Azure Industrial IoT OPC UA integration is overkill or simply a step later in the project plan. Enjoy!   End to end Integration of Kepware to ThingWorx using MQTT over Azure IoT (YoutTube 45 minute deep-dive)   ThingWorx entities for import (ThingWorx 9.0)   This approach can be quite good for a simple demo if you have a Kepware Integrator or Kepware Enterprise license, but the use of IoT Gateway for many servers and tags can be quite costly.   Those looking to leverage Azure IoT Hub for MQTT integration to ThingWorx would likely also find this recorded session and shared utilities quite helpful.   Cheers, Greg

Recently I have been accompanying an integration partner and end customer around an issue experienced with ThingWorx resource exhaustion.  Early on it seemed like this was an issue with the ThingWorx Azure IoT Hub Connector as it would freeze up and become unresponsive.  Following a root cause analysis it became clear that it was actually caused by a lack of a number of standard cloud design patterns, which if used would have automatically adapted operation of the overall solution to be far more resilient as well as resource optimized.   The way that the logic was structured, it prioritized job execution on entities with the oldest last success time and would continue to retry these executions (IoT Direct Methods) every few seconds until successful.  There were a number of problems here, but I'll unpack a few in order to tie the problem to the solution via design patterns.   1) No exception handling When the direct method execution failed/timed out or the system reported being unable to execute the remote service, this response was not used to adapt the solutions behavior. 2) No backoff retry mechanism As exceptions were not caught, an adaptive retry mechanism with incremental or exponential backoff could not be leveraged to limit the impact of the build up of the failing retries. 3) No exception tracking Tracking that exceptions were occurring and counting them would allow powering an exponential backoff retry algorithm (with jitter), a Cancel or Circuit Breaker pattern (stop doing something which is just broken), as well as provided alerting to address specific areas of the distributed solution experiencing issues. 4) Conflicting priorities It was interesting to see the manifestation of the conflicting interests of wanting to ensure checks and balances (had all needed data) and system resiliency.  Retries and resource usage built up exponentially due to the transient error instead of backing them off.  Trying so hard to get the needed data from failing sensors meant that operational sensors were deprioritized and their data was not received either - spreading the localized issue to the whole system.   Around the time that I shared my recommendations and some examples of how to make the solution more resilient, one of my technical colleagues at Microsoft shared some extremely interesting and relevant design patterns documented by Microsoft as a part of the "Microsoft Azure Well-Architected Framework".  This framework with included Design Patterns for specific cloud application goals allows applying well-known industry standard approaches to dealing with the challenges of large scale distributed enterprise systems (reliability, performance, cost optimization).   She later then shared this blog post describing exactly the exponential backoff retry with jitter pattern which we had together recommended to the systems integrator.   What's interesting for us ThingWorx people is that this framework from Microsoft is about well-architected cloud solutions and does not specifically reference the Azure stack, and as such many of these approaches and design practices can be employed in your ThingWorx applications.  What are you waiting for?  Go check them out!

Hi Community,   Although we have reference architectures and integration paths for connecting devices to ThingWorx through Azure IoT; no one has ever written anything about doing the same from one ThingWorx to another.  I thought I’d change that and put some ideas out there around how one might go about doing this.  Although this is not officially supported or recommended by PTC; I have consulted with a number of leading SMEs on the subject, which have participated in forming the basis of my thinking outlined here.   Components Required (in order of communication path): On-premise ThingWorx Platform Protocol Adapter Toolkit* (CXS) - MQTT Azure IoT Edge Azure IoT Hub ThingWorx Azure IoT Hub Connector (CXS) Azure Cloud-hosted ThingWorx Platform   PAT (2) with codec to encode MQTT messages publishes to on-premise IoT Edge MQTT endpoint which handles store-and-forward of messages to IoT Hub.  An Azure IoT device would exist for each Thing you wish to represent on the ThingWorx servers.  The Azure IoT Hub Connector would pick-up the incoming messages and pass them on to the cloud ThingWorx which would decode the MQTT payload and map to Thing property updates.   The only part that I presently don’t like about this approach is that you’ll need to decode the MQTT messages on the ThingWorx platform in the cloud when they are received from the IoT Hub, and this mechanism will need to also need to handle encoding and publishing back to the IoT Hub if C2D (Cloud-to-Device) messages are to be implemented (aka bi-directional).  This is required as ThingWorx only supports AlwaysOn as an application level protocol so some form of mapping needs to be done.   * Another approach would be to replace the PAT with a custom agent which implements both the ThingWorx Edge SDK and the Azure IoT device SDK   Regards,   Greg Eva

Here is a spreadsheet that I created which helps to estimate data transfer volumes for the purpose of estimating egress costs when transferring data out of region.   You find that there are a number of input parameters like numbers of assets, properties, file sizes, compression ratio, as well as a page with the cost elements which can be updated from the Interweb.    

ThingWorx Foundation Flow Enable customers using Azure to take advantage of Azure services Access hundreds of Azure system connectors by invoking Azure Logic Apps from within ThingWorx Flow Execute Azure functions to leverage Azure dynamic, serverless scaling and pay just for processing power needed Access Azure Cognitive AI services for image recognition, text to voice/voice to text, OCR and more Easily integrate with homegrown and commercial solutions based on SQL databases where explicit APIs or REST services are not exposed Automatically trigger business process flows by subscribing to Windchill object class and instance events Provide visibility to mature PLM content (such as when a part is released) to downstream manufacturing and supply chain roles and systems Easily add new actions by extending functionality from existing connectors to create new actions to facilitate common tasks Inherit or copy functionality from existing actions and change only what is necessary to support new custom action Azure Connector SQL Database Connector Windchill Event Trigger Custom Action Improvements Platform Composer: Horizontal tab navigation is back!  Also new Scheduler editor. Security: TLS 1.2 support by default, new services for handling expired device connections New support for InFlux 1.7 and MSSQL 2017 * New* Solution Central Package, publish and upload your app with version info and metadata to your tenancy of Solution Central in the PTC cloud Identify missing dependencies via automatic dependency management to ensure your application is packaged with everything required for it to run on the target environments Garner enterprise-wide visibility of your ThingWorx apps deployed across the enterprise via a cloud portal showcasing your company’s available apps, their versions and target environments to foster a holistic view of your entire IIoT footprint across all of your servers, sites and use cases Solution Central is a brand-new cloud-based service to help enterprises package, store, deploy and manage their ThingWorx apps Accelerate your application deployment Initially targeted at developers and admins in its first release, Solution Central enables you to: Mashup Builder 9 new widgets, 5 new functions. Theme Editor with swappable Mashup Preview Responsive Layout enhancements including new settings for fixed and range sizes New Builder for custom screen sizes, new Widget and Style editors, Canvas Zoom Migration utility available for legacy applications to help move to latest features Security 3 new built-in services for WebSocket Communications Subsystem: QueryEndpointSessions, GetBoundThingsForEndpoint, and CloseEndpointSessions Provide greater awareness of Things bound to the platform Allow for mass termination of connections, if necessary Can be configured to automatically disconnect devices with expired authentication methods Encrypting data-in-motion (using TLS 1.2) is a best practice for securely using ThingWorx For previous versions, the installer defaulted to not configuring TLS; ThingWorx 8.5 and later installers will default to configuring TLS ThingWorx will still allow customers to decline to do so, if desired Device connection monitoring & security TLS by default when using installer   ThingWorx Analytics Confidence Model Training and Scoring (ThingWorx Analytics APIs) Deepens functionality by enabling training and scoring of confidence models to provide information about the uncertainty in a prediction to facilitate human and automated decision making Range Property Transform and Descriptive Service Improves ease of implementation of data transformations required for common statistical process control visualizations Architecture Simplification Improves cost of ownership by reducing the number of microservices required by Analytics Server to reduce deployment complexity Simplified installation process enables system administrators to integrate ThingWorx Analytics Server with either (or both) ThingWorx Foundation 8.5 and FactoryTalk Analytics DataFlowML 3.0.   ThingWorx Manufacturing and Service Apps & Operator Advisor Manufacturing common layer extension - now bundling all apps as one extension (Operator Advisor, Asset Advisor, Production KPIs, Controls Advisor) Operator Advisor user interface for work instruction delivery Shift and Crew data model & user interface Enhancements to Operator Advisor MPMLink connector Flexible KPI calculations Multiple context support for assets   ThingWorx Navigate New Change Management App, first in the Contribute series, allows a user to participate in change request reviews delivered through a task list called “My Tasks” BETA Release of intelligent, reusable components that will dramatically increase the speed of custom App development Improvements to existing View Apps Updated, re-usable 3D viewing component (ThingView widget) Support for Windchill Distributed Vaults Display of Security Labels & Values   ThingWorx Azure IOT Hub Connector Seamless compatibility of Azure devices with ThingWorx accelerators like Asset Advisor and custom applications developed using Mashup Builder. Ability to update software and firmware remotely using ready-built Software Content Management via “ThingWorx Azure Software Content Management” Module on Azure IoT Edge. Quick installation and configuration of ThingWorx Azure IoT Hub Connector, Azure IoT Hub and Azure IoT Edge SCM module.   Documentation ThingWorx Platform ThingWorx Platform 8.5 Release Notes ThingWorx Platform Help Center ThingWorx 8.5 Platform Reference Documents ThingWorx Connection Services Help Center   ThingWorx Azure IoT Hub Connector ThingWorx Azure IoT Hub Connector Help Center   ThingWorx Analytics ThingWorx Platform Analytics 8.5.0 Release Notes Analytics Server 8.5.1 Release Notes ThingWorx Analytics Help Center   ThingWorx Manufacturing & Service Apps and ThingWorx Operator Advisor ThingWorx Apps Help Center ThingWorx Operator Advisor Help Center   ThingWorx Navigate ThingWorx Navigate 8.5 Release Notes Installing ThingWorx Navigate 8.5 Upgrading to ThingWorx Navigate 8.5 ThingWorx Navigate 8.5 Tasks and Tailoring Customizing ThingWorx Navigate 8.5 PTC Windchill Extension Guide 1.12.x ThingWorx Navigate 8.5 Product Compatibility Matrix ThingWorx Navigate 8.5 Upgrade Support Matrix ThingWorx Navigate Help Center     Additional Information Helpcenter ThingWorx eSupport Portal ThingWorx Developer Portal PTC Marketplace The National Instruments Connector can be found on PTC Marketplace  

  Use the C SDK to build an app that connects to ThingWorx with persistent bi-directional communication   Guide Concept This project will introduce more complex aspects of the ThingWorx C SDK and help you to get started with development.  Following the steps in this this guide, you will be ready to develop your own IoT application with the ThingWorx C SDK.  We will teach you how to use the C programming language to connect and build IoT applications to be used with the ThingWorx Platform.   You'll learn how to Establish and manage a secure connection with a ThingWorx server, including SSL negotiation and connection maintenance Enable easy programmatic interaction with the Properties, Services, and Events that are exposed by Entities running on a ThingWorx server Create applications that can be directly used with your device running the C programming language Basic concepts of the C Edge SDK How to use the C Edge API to build a real-world application How to utilize resources provided in the Edge SDK to help create your own application NOTE: This guide's content aligns with ThingWorx 9.3. The estimated time to complete ALL 3 parts of this guide is 60 minutes.   Step 1: Completed Examples Download the completed files for this tutorial: ThingWorx C Edge SDK Sample Files.zip.  This tutorial will guide you through working with the C SDK on differing levels. Utilize this file to see a finished example and return to it as a reference if you become stuck creating your own fully fleshed out application.  Keep in mind, this download uses the exact names for Entities used in this tutorial. If you would like to import this example and also create Entities on your own, change the names of the Entities you create.   Step 2: Environment Setup In order to compile C code, you need a C compiler and the ThingWorx C Edge SDK available in the PTC Support download site.  It will be helpful to have CMake installed on your system. CMake is a build tool that will generate make or project files for many different platforms and IDEs.   Operating System      Notes Windows You will need a 3rd party compiler such as MinGW GCC, Cygwin GCC or you can follow these Microsoft instructions to download and use the Microsoft Visual C++ Build Tool. Mac Download the Apple Developer Tools. Linux/Ubuntu A compiler is included by default.   NOTE: You can use CMake, version 2.6.1 or later to build projects or make files, which then are used to build the applications that you develop with the C SDK.     Before you can begin developing with the ThingWorx C SDK, you need to generate an Application Key and modify the source code file. You can use the Create an Application Key guide as a reference.   Modify Source File Extract the files from the C SDK samples zip file. At the top level of the extracted files, you will see a folder called examples. This directory provides examples of how to utilize the C SDK. Open a terminal, go to your workspace, and create a new directory. You can also just switch to the unzipped directory in your system. After you've created this directory in your workspace, copy the downloaded files and folders into your new directory. You can start creating your connection code or open the main.c source file in the examples\SteamSensor\src directory for an example. Operating System      Code Linux/Ubuntu gedit main.c OR vi main.c Mac open –e main.c Windows start main.c Modify the Server Details section at the top with the IP address for your ThingWorx platform instance and the Application Key you would like to use. Change the TW_HOST definition accordingly. Change the TW_PORT definition accordingly. Change the TW_APP_KEY definition to the keyId value saved from the last step.         /* Server Details */ #define TW_HOST "https://pp-XXXXXXXXX.devportal.ptc.i" #define TW_PORT 80 #define TW_APP_KEY "e1d78abf-cfd2-47a6-92b7-37ddc6dd34618"​         NOTE: Using the Application Key for the default Administrator is not recommended. If administrative access is absolutely necessary, create a User and place the user as a member of Admins.   Compile and Run Code To test your connection, you will only need to update the main.c in the SteamSensor example folder. CMake can generate Visual Studio projects, make build files or even target IDEs such as Eclipse, or XCode. CMake generates a general description into a build for your specific toolchain or IDE.   Inside the specific example folder you would like to run, ie SteamSensor. Create a directory to build in, for this example call it bin. mkdir bin  cd bin Run the CMake command listed below. This assumes CMake is already on your PATH.         cmake ..​           CMake has now produced a set of project files which should be compatible with your development environment. Operating System        Command                                            Notes Unix make A set of make files Windows msbuild tw-c-sdk.sln /t:build A visual studio solution   NOTE: CMake does its best to determine what version of Visual Studio you have but you may wish to specify which version to use if you have more than one installed on your computer. Below is an example of forcing CMake to use a specific version of Visual Studio: cmake -G "Visual Studio 15 2017" .. If your version of Visual Studio or other IDE is unknown, use cmake -G to see a list of supported IDEs.   You also have the alternative of opening the tw-c-sdk.sln from within Visual Studio and building in this IDE.   NOTE: By default, CMake will generate a build for the creation of a release binary. If you want to generate a debug build, use the command:           cmake -DBUILD_DEBUG=ON ..           Once your build completes you will find the build products in the CMake directory (see example below). From here, open the project in your IDE of choice.   NOTE: You should receive messages confirming successful binding, authentication, and connection after the main.c file edits have been made.   Operating System Files Description Unix ./bin/src/libtwCSdk_static.a  Static Library Unix ./bin/src/libtwCSdk.so  Shared Library Unix ./bin/examples/SteamSensor/SteamSensor   Sample Application Windows .\bin\src\<Debug/Release>\twCSdk_static.lib  Static Library Windows .\bin\src\<Debug/Release>\twCSdk.dll  Shared Library Windows .\bin\examples\<Debug/Release>\SteamSensor\SteamSensor.exe  Sample Application   Click here to view Part 2 of this guide.