IIoT, Industrie 4.0

AI-driven insights at the industrial edge

Edge technology such as smart servers can be leveraged create a uniform fabric of data and automation that ties together control systems, Internet of Things (IoT) devices and artificial intelligence (AI) engines. Case study shows up to 30% reduction in maintenance and call out costs and as much as 30% energy savings.

By Dialog Semiconductor May 26, 2021
Courtesy: Dialog Semiconductor

 

Learning Objectives

  • Internet of Things (IoT) and artificial intelligence (AI) innovation help companies gain faster and more actionable insights./li>
  • Cost and complexity are the biggest challenges when picking an edge and AI platform.
  • Benefits of leveraging an edge server combined with a cloud-based AI platform, include better business outcomes and more powerful monitoring and control.

Industry 4.0 and the Industrial Internet of Things (IIoT) has put an unprecedented demand for data that exists in control systems as more and more organizations across various industries are realizing its power to allow more nuanced and transformative decision making for operations.

Data from machine-to-machine (M2M) systems can augment IoT- enabled sensors and cameras to identify patterns of system operations that can be intelligently optimized through the use of artificial intelligence (AI) models. However, it’s important to be mindful there are still significant challenges around how to access, decipher and transform the massive amounts of data being collected by these M2M systems, as the customer application case study below demonstrates.

Organizations can leverage edge technology in several ways, including smart servers, to access and create a uniform fabric of data and automation that ties together control systems, IoT devices and AI engines.

Overcoming IIoT challenges

Organizations can leverage their data to glean useful information about what is happening within their systems. Analyses can reveal patterns within the data, creating valuable insights that help companies simplify operational workflows and maximize system uptime while cutting maintenance costs and energy consumption.

Consider Airedale International Air Conditioning, a designer and manufacturer of precision thermal management systems. The systems, which area designed for maximum efficiency and minimal environmental impact, employ sophisticated monitoring and control electronics. They recognized the potential of AI to transform the operational efficiency of HVAC systems. They are leveraging AI, specifically in the form of machine learning (ML), to enable vast volumes of system data to be classified, grouped and analyzed.

The challenge, which is common among many IIoT applications, is collating diverse, multi-protocol network data from the edge, close to sensors and actuators, send it to ML platforms hosted in the cloud and then communicate and deploy the results of the processed data. Add this to the complexity of ensuring top-notch data security, which is paramount throughout the process.

Airedale has a large installed base of equipment in data centers and is diversifying into other sectors. To succeed, the company needed to find a flexible, scalable implementation that could be deployed in almost any industrial or commercial environment. In doing this, it would be able to optimize its offering to existing customers, all while creating incremental sales, and expanding its potential market to encompass other industry sections.

Providing secure, scalable technologies

When looking for a solution, there are a number of factors to take into account, but the two biggest are cost and complexity. After reviewing the options, Airedale chose to leverage a multi-protocol IoT edge server and cloud-based AI platform.

The open IoT edge server selected solves the complexities of integrating legacy technologies with innovations in cloud analytics and AI. This means the edge and cloud AI can co-exist securely with transitional automation networks such as building automation systems (BAS) using web service application program interfaces (APIs). This enhances ROI for facilities that have invested in BAS systems and want to enhance it with AI.

Out-of-the-box device drivers, controls and automation services, intuitive management system and easy-to-use programming tools enables custom applications development and rapid field deployment expediting insights from data, providing safer operations, and creating greater efficiencies and operational cost savings. When paired with the AI platform, the IoT edge server can co-exist with existing operational or building management systems (BMS), enhancing both by creating a feedback loop that ties them together.

Airedale chillers have integrated programmable logic controllers (PLCs), which connect the company’s building management system (BMS) using IP, Modbus RS485, or BACnet protocols. A further two-way Modbus RS485 connection from each controller feeds the edge server’s IoT unit. Using an open API, the edge server then interfaces to the cloud platform using the message queuing telemetry transport (MQTT) protocol running over a 4G cellular radio link.

Data calls and control signals are routed back to the PLCs in the chillers via a virtual private network (VPN) client, which runs on the edge server’s IoT. As a result, customers can access analytics and reporting online via the cloud-based platform. This is all done over a secure data pipeline, and the development of control sequences and custom visualization is made easy and intuitive thanks to the edge server’s drag-and-drop programming tools.

When paired with Smartia’s AI platform, Dialog Semiconductor’s SmartServer can co-exist with existing operational or building management systems (BMS), enhancing both by creating a feedback loop that ties them together. Using Dialog Semiconductor’s IoT Access Protocol (IAP) open API, the SmartServer interfaces to the Smartia MAIO cloud platform using the MQTT protocol running over a 4G cellular radio link. Data calls and control signals are routed back to the PLCs in the chillers via Smartia’s VPN client, which runs on the SmartServer IoT. Courtesy: Dialog Semiconductor

When paired with Smartia’s AI platform, Dialog Semiconductor’s SmartServer can co-exist with existing operational or building management systems (BMS), enhancing both by creating a feedback loop that ties them together. Using Dialog Semiconductor’s IoT Access Protocol (IAP) open API, the SmartServer interfaces to the Smartia MAIO cloud platform using the MQTT protocol running over a 4G cellular radio link. Data calls and control signals are routed back to the PLCs in the chillers via Smartia’s VPN client, which runs on the SmartServer IoT. Courtesy: Dialog Semiconductor

Realizing the benefits of the edge

The benefits of leveraging an edge server combined with a cloud-based AI platform, include better business outcomes and more powerful monitoring and control for customers.

For Airedale, the edge server enabled the company to diversify its business through the openness, flexibility and scalability of the technology. It offers open, multi-channel support for connecting to BACnet, IAP, LON and Modbus networks, within legacy installations or in new designs. The edge server harmonizes data from these networks, simplifying the process of linking them to the AI platform.

Businesses strive to improve customer experiences and part of this is giving them more powerful monitoring and control. Industrial systems are full of information related to data correlations, trends, anomalies, deviations, root causes and other key operational information. By analyzing and presenting data for actionable use, their customers can improve business decisions.

Beyond the overall improved effectiveness of the HVAC systems, the company’s customers also will see reduced maintenance costs and energy savings.

Algorithms, machine-learning techniques

Using a variety of algorithms and machine learning techniques, the system can conduct an ongoing performance analysis. Take a refrigerant-based cooling system as an example. The system can measure superheat, sub-cool, suction and heat pressures, water and air flow. It analyzes these factors for deviations in normalized behavior while considering the installation’s overall duty cycle, efficiency and power consumption. If it detects a change or drop in performance, customers will be notified so a maintenance team can step in. This level of early intervention can result in up to 30% reduction in maintenance and call out costs. By stepping in early, organizations can prevent prolonged periods of higher energy consumption and eventual breakdown.

The HVAC unit’s load profile also will be analyzed to determine if it is possible to improve operating efficiency which can result in energy savings from 0 to 30%. For example, if an air handling unit (AHU) serves a meeting room only used at certain times of the day, the system would detect and let the user know it is wasting energy by being left on all day. Then, the system will autonomously learn to correct this behavior and turn the air handling unit (AHU) off or adjust the temperature and airflow setpoint as needed.

The industrial world is one that, historically, has been set in its ways. However, many are now looking to accelerate their digital transformation to keep up with Industry 4.0. To stay ahead of the competition, businesses are looking to deliver tangible benefits for customers and are leveraging intelligent technology to do so. With the help of innovations in AI and IoT, organizations can deliver customers more actionable insights to improve business.

Dialog Semiconductor article edited by Chris Vavra, web content manager, Control Engineering, CFE Media & Technology, cvavra@cfemedia.com.

CONSIDER THIS

What benefits can you derive from edge-based artificial intelligence (AI) solutions?


Dialog Semiconductor