Data Collection, Standardization, and Organization 

The system uses AUTOSOL ACM to connect to thousands of distributed field devices using protocols like Modbus, ABB Totalflow, and Allen-Bradley Ethernet IP. ACM then hosts the tags for Ignition to read over OPC-UA. Ignition uses UDTs and a strict tag hierarchy to maintain standardization between different equipment of the same type. 

Ignition collects data, stores trends in the tag historian, and displays tag data on dynamic screens. Navigation and displays are driven by database tables to automatically update when new sites are added. Screens use high-performance HMI design to give operators at-a-glance equipment health and performance information. 

Automated Equipment Rollout Process 

DMC designed the system to automatically generate all required equipment configuration from a template spreadsheet file. This creates AUTOSOL import files and Ignition tags, so users do not have to manually add equipment to the system. The auto-generation process is designed to both create new equipment and update existing equipment, making it easy to roll out changes to the field.  

Data Reporting and Integrations 

DMC developed custom dashboards and reports to display critical production information. Reports contain a mix of user-submitted information, trend aggregations, and live data. Reports and data can be accessed via several different means: 

• Ignition dashboard screens 
• Export-to-CSV downloads 
• REST API data endpoints 
• Alarm text, email, and phone callouts 

Conclusion 

With their new Ignition/AUTOSOL SCADA system, the client has much greater visibility into their field-wide data. They can build on DMC’s modular architecture to quickly roll out new sites and develop new features.  

Learn more about DMC’s Ignition SCADA Programming expertise and contact us for your next project.

The post Ignition and AUTOSOL Distributed Oil & Gas Cloud SCADA appeared first on DMC, Inc..

DMC first audited and updated the plant’s interlock. During this process, DMC found several interlocks needed to be bypassed for normal operations and the majority did not serve adequate safety or protection purposes. DMC leveraged the built-in functionality of the PCS7 APL Interlock objects to link devices together so that operators could click on and navigate through the whole interlock matrix to see what was stopping them and why they were interlocked. This added clarity, reliability, safety, and insight into the system, so that the client could now run manually, if needed. 

Once the batch rewrite in PCS7 was complete, DMC implemented a historian and batch reporting system in Ignition to provide better visibility into the plant’s processes. We used OPC UA to pull data via the PCS7 WinCC 7 HMI using the PCS7 Connectivity Pack. 

After experiencing the capabilities of Ignition, the client decided to add additional features beyond data collection and reporting. One of the client’s goals was to streamline their data and document handling surrounding material shipping and receiving, which was previously managed through multiple manually maintained spreadsheets. DMC developed a custom integration to the client’s LIMS systems to compile batch quality test data from multiple sources into a single database. DMC then developed integrations to PCS7 and the client’s truck loadout system to automatically track and release batches for loadout and generate BOLs incorporating the batch quality data. In the end, the entire loadout process could be carried out on a tablet, including operator interfaces for loadout inspection and signoff signatures. 

Learn more about DMC’s Manufacturing Automation and Intelligence expertise and contact us today for your next project. 

The post S88 Batch Architecture Implementation Using PCS7 appeared first on DMC, Inc..

• Send a message to the IoT hub (device to cloud messaging)
• Receive messages from the IoT hub (cloud to device messaging)

The customer was seeking to improve their operations by connecting their field assets to the cloud. In the field, NI CompactRIO (cRIO) controllers were running a custom LabVIEW real-time app to control hardware and log data on the field equipment. The data was all there, but the data was distributed and only accessible from HMIs out in the field. They needed a simple solution to make their data actionable and viewable centrally.

The web app’s APIs were available to the LabVIEW app via LabVIEW’s HTTP Client toolkit. When the LabVIEW app collected data, it simply made an HTTP request to the web app to send the message to the IoT hub. This limited changes and impact on the LabVIEW real-time app.

Our customer’s equipment is often in remote locations, so internet connectivity is not consistent. DMC built in a local caching implementation using SQLite. This allowed offline field devices to preserve messages without internet connection. With caching, the web app could also build larger messages, limiting the number of messages sent to the IoT hub. Azure IoT hub charges per message, so this is a cost-saving feature.

The connection to Azure provided our customer with valuable insights into their field equipment. They can better schedule with live data, analyze failures, and perform preventative maintenance. The Azure IoT Hub provides a secure and scalable platform for aggregating and storing data and the web app. The Azure IoT Hub also enables many data analysis and visualization tools such as CosmosDB and PowerBI dashboards.

Learn more about DMC’s LabVIEW Programming and contact us today for your next project.

The post NI-cRIO Connection to Azure IoT Hub appeared first on DMC, Inc..

The NFPA Publication 70, NEC, and CEC define Class II locations as those in which combustible dust may be found. The subcategory of divisions that the classes are further subdivided into defines the likelihood of the hazardous material being present in a flammable concentration.

Division 2 is defined as follows:

“In which ignitable concentrations of hazards are handled, processed, or used, but which are normally in closed containers or closed systems from which they can only escape through accidental rupture or breakdown of such containers or systems.

We used a Factory Acceptance Test (FAT) because the hazardous environment was not present. This FAT tested functionality and verified that, if someone were to open the door, the purge would be lost, and that we killed power to the entire panel. Overall, the test proves that we eliminated the risk.

Our engineers provided support in successfully completing this project the client had already started. With our expertise and programming experience, we successfully provided a solution that can operate in a C1D2 hazardous environment.

Read more about DMC’s manufacturing automation & intelligence services and contact us to get started on your next project.

The post Control Panel Design for Hazardous Locations appeared first on DMC, Inc..

Part 1: Unsupervised Learning – Pattern Search

DMC preprocessed the data using K-Nearest Neighbor’s (KNN) imputation, feature scaling, normalization, and Principal Component Analysis (PCA). DMC then created clustering models to discover the underlying pattern in the raw data and decoded the pattern to produce meaningful data. Later, evaluation matrices (Silhouette Score and Davies Bouldin Score) and cluster visualization techniques (T-Distributed Stochastic Neighbor Embedding and Principal Component Analysis) were applied to evaluate the results.

Oil wells that differed in condition (location, depth, equipment, etc.) had different telemetry performances. By clustering the well condition data, we could identify if there were similarities between certain well conditions and if clusters were substantially different from each other. Analyzing the corresponding telemetry performance of the clusters guided us to set up wells with better telemetry performance.

Part 2: Supervised Learning – Diagnosis and Prediction

After Phase I (Unsupervised Learning), DMC created a more powerful machine learning model capable of diagnosing well conditions and providing guidance on ways to optimize telemetry performance. We designed the model to predict telemetry performance for newly-acquired sets of well condition data.

To train supervised learning models, the prediction goal (telemetry performance) must match up with the inputs (well conditions). The raw telemetry performance data consists of time traces, so to prepare the training data, DMC extracted features from the time traces and paired them up with corresponding well condition data using Dataiku.

With the training data well prepared, DMC preprocessed the dataset using one-hot encoding, imputation on the missing values, and feature scaling. DMC then trained non-parametric models (K-Nearest Neighbors and Decision Tree Regression) and parametric models (Lasso regression, Kernel Ridge Regression) as baselines. Finally, a Deep Neural Network was developed to perform the diagnosis and prediction tasks.

In order to use the neural network to improve telemetry performance, DMC designed feature importance analysis methods targeting specific diagnostics. DMC used several statistical approaches to identify numeric and categorical features, then ranked the conditions of a well to optimize telemetry performance.

Part 3: Telemetry Analytics Dashboard – Visualization

The telemetry analytics web interface was designed using Plotly for the purpose of visualizing the massive dataset, running statistical analysis, and displaying the resulting graphs. The dashboard provided a tool for the client to easily visualize the data and obtain information without being exposed to implementation details.

The dashboard displayed the oil wells on a map based on their recorded geographic locations, allowing users to click on a well location to reveal detailed information about the well. Users could select multiple wells on the map or use filters to select wells meeting certain criteria. Based on the selected data, the dashboard can run statistical analyses and display a variety of visualizations (heat maps, word clouds, histograms). Users can also export selected data and graphs and save filter configurations for later use.

Learn more about DMC’s Test and Measurement expertise and contact us for your next project. 

The post Machine Learning and Telemetry Analytics appeared first on DMC, Inc..

To ensure these drill bits are consistently up to standard, they needed to be inspected both before and after operation. Prior to DMC’s help, this inspection process was lengthy, inefficient, and required a lot of manual input. Therefore, our client reached out to DMC to better automate the inspection process.

Subsystems

DMC’s development effort for this project can be broken down into two main components. The first subsystem DMC developed was a device-routing system controlled by a PLC to load and unload the drill bits into the inspection cells. This subsystem included a REST API to an orchestrator previously developed by the client that managed overall system sequencing. This API interfaced with the PLC to control servo drives to route and position drill bits accurately and efficiently.  

The second subsystem DMC developed were cells to carry out the inspection operations. Each inspection cell consisted of a six-degree-of-freedom (DOF) UR10 robot arm and a rotary table, which is able to independently rotate the drill bit under inspection. The additional degree of freedom afforded by the rotary table allows for a greater functional workspace of the robot arm and therefore greater flexibility in inspection operations.

Microservice Development

To coordinate the motion of the redundant DOF with the UR10 and perform inspection operations, DMC developed a Docker-based microservice. This microservice provided a REST API developed in collaboration with the client and exposed a set of operations to their orchestrator. This API included a variety of endpoints—most critically, a set of commands to load and execute inspection operations including point-to-point and scan-path motions.

The microservice then interfaced with Energid’s Actin SDK to generate specific joint commands to each hardware device. These commands were then relayed through Energid’s UR hardware plugin as well as a custom-developed, MQTT-based Actin hardware plugin to the rotary table. By wrapping and containerizing the Actin SDK’s motion control logic, the system was able to separate the inspection operations from hardware control and business logic. Inspection actions could then be triggered from simple REST endpoints without requiring any knowledge of the path planner or hardware control.

Image Analysis

Within the inspection cell, cameras take RGB images of the drill bit at each orientation. These images are compiled and analyzed to detect damage at each of the around forty cutter heads. A laser scanner also generates a three-dimensional point cloud of the drill bit, which is used for analysis as well.

DMC’s solution not only made the inspection process more efficient and cost-effective, but also reliable and sustainable. By containerizing the inspection cell operations, DMC was able to work with the end client to develop a set of functional tests which could be executed on the inspection cell software by executing REST API calls in sequence. Additionally, by separating the hardware layer out as a set of well-defined communication interfaces to the control logic container, DMC was able to create virtual hardware which could be actuated from the container. Automated testing was further developed to integrate these functional tests with build pipelines including Jenkins.

Read more about DMC’s PLC Programming expertise and contact us to get started on your next project.

The post Automated Inspection System in the Oil and Gas Industry appeared first on DMC, Inc..

Monitoring Distributed Systems

Having previously switched monitoring services for their cogenerators several times, the client had around 250 cogenerators in the field running on different platforms. This meant that monitoring crews had to collect data from disparate apps by navigating through a tangle of user interfaces and systems for various portions of their business operations.

WinCC OA was selected as the new control system to introduce a more coherent system for tracking and maintenance. This also provided the possibility to unite both future and existing systems under a single umbrella. Given the scale of the operations, WinCC OA’s native support of distributed systems was well-suited to handle data over a large geographic range in a heavily-customized fashion. DMC’s goals included facilitating flexible setup and management of hundreds of sites in bespoke configurations across state lines.

Initially, DMC created an example interface and configuration for a single type of cogenerator generator. To do this, we developed a system to configure a new PLC and generator through a simplified menu that dynamically adds a new cogenerator to a site and populates its data. This reduced the process of setting up monitoring for a new generator to just a dozen mouse clicks and allowed the customer to easily deploy the system to arbitrarily large counts onsite and arbitrarily large site counts.

Monitoring Load Modules

Building on these successes, DMC extended the cogenerator monitoring framework to the other core component of the client’s cogeneration systems – the load modules. DMC worked with the client to generate a standardized set of definitions for hot water loads such as domestic hot water, space heating, and others, built on a repeatable set of pumps, valves, heat exchangers, and other components  These new modules and the existing cogenerator modules were then aggregated onto a configurable site diagram with itemized representations, pass-through of legacy overviews, and a user-customized overview. The fully WinCC OA-based system allows customizable configuration of the cogenerators and the loads with a clear path forward for new device and module types. Benefits of this system include rapid deployment, rapid load type definition and development, automated data display for the entire site, and fully-integrated generator and load control.

Employee Training

As part of this effort, the client’s employees trained with DMC to gain skills to manage their own development within this framework. This enabled the client to increase responsiveness to new initiatives or requirements and extend the existing architecture to any incremental device or design changes. They were also better equipped to support and design for further integrations, including upcoming work on new cogenerator types and site configurations. They were able to gain confidence in their knowledge and DMC’s support when it comes to approaching customization and maintenance of the system.

Future Developments

Due to the level of success with this project, DMC continues to support high-level developments on this system. Current upgrades include more configurable options on cogenerator interfaces, improvements to set-up templates, and the integration of new varieties of loads, site configurations, and custom PLC drivers with WinCC OA. We’ll be integrating some of the styling from the WinCC OA Open Library developed by DMC.  

Learn more about DMC’s partnership with Siemens.

The post Distributed Cogenerator Control System appeared first on DMC, Inc..

DMC programmed a Siemens 1500 PLC using the “Optional Hardware Configuration” feature discussed in this blog. This allowed users to reconfigure the PROFINET IO network according to a predefined topology. As new ET200 racks were added to the network, the PLC would automatically assign PROFINET names, IP Addresses, and IO configuration.

DMC also leveraged the full feature set of WinCC Advanced in order to expose all relevant process information to the operators. This included a status log that contained all batch events, a custom text log for each material, an alarm log for delivery faults, and data trending to graphically show hopper weights over the past several days. This data could be both viewed on the touch panel HMI screen and exported for further analysis.

DMC upgraded the client’s SCADA program, developed with WinCC 7, to interface with the new delivery system. Operators are now able to view unit status and delivery quantities through an industrial PC and can initiate material delivery as part of a larger batch.

Learn more about DMC’s PLC Programming and HMI Programming services.

The post Expandable Ingredient Dosing System Using Siemens PLC appeared first on DMC, Inc..

After understanding the client’s needs and identifying the shortcomings of the existing system, DMC worked with the client to establish a standardized hub and spoke architecture which will be implemented at all new facilities and retroactively at existing facilities. This architecture is designed from the ground up to avoid any loss of compliance data.  As shown in Figure 1 below, each site will have a remote Ignition gateway server. These remote gateways can store data the event of a communication loss. All of that data is seamlessly transferred to the central Ignition server once communication is restored.

Figure 1: Ignition Hub & Spoke Architecture

The client was very satisfied, and DMC will continue to provide ongoing support and new feature requests for their system. 

Learn more about our Ignition Designer Programming expertise. Contact us for more information.

The post Ignition Programming for a Geographically Distributed Renewable Energy Firm appeared first on DMC, Inc..

DMC developed control code capable of dynamically configuring the PLC and HMI at runtime based on the number of wells at the site. Based on the site’s configuration, the program:

• Updates the PLC hardware configuration to include the appropriate number of I/O cards. This allows the client to only install the required hardware without causing hardware faults.
• Configures the HMI to only display wells that were drilled, preventing operator confusion with extraneous HMI data.
• Alters the program execution to only execute logic for the drilled wells. This improved scan time, eliminated “phantom” alarms, and simplified troubleshooting.

By leveraging the industry’s best practices in object-oriented PLC programming, DMC built one robust solution that the customer can use repeatedly on a wide mix of well sites. The flexibility of this system will allow the client to expand well sites in the future without costly developer involvement. Further, the uniformity between assets allows for easier operator training and troubleshooting.

Learn more about DMC’s expertise in PLC programming and the Oil and Gas Industry.

The post Oil Well Control System Upgrade using an Allen-Bradley Control Logix PLC appeared first on DMC, Inc..