What's New
Tech News
ITG Delivers Real Value with Machine Learning Applications
Artificial intelligence (AI) as a single force is driving the world’s economy in ways not seen in a long time. Information touting the potential of the field has flooded the media and created a fervor amongst the general population and with investors. Major markets report that the growth seen in 2023 has been fueled by AI and led by key players in the field. Sources indicate that more than 35% of businesses are using AI and sophisticated companies routinely have AI or its functional subset machine learning (ML) in the top three of current initiatives.
It should be no surprise that the industrial space offers rich opportunities to apply this technology. Factories vary in the mix of product-specific equipment, but it is difficult to find a facility that doesn’t include systems for managing the environmental properties of areas or processes within. Where consumer applications may rely on inputs from humans and other complex derivatives, modern industrial systems are made of connected components which can provide reliable, consistent information to a modeling system and easily respond based on commands from a learned controller.
ITG Technologies (ITG) has been actively helping clients in this space since 2014. The company has existed for more than twenty years, and they feature a deep and diverse background in the general fields of factory and industrial automation. That experience, when coupled with the SORBA’s IIoT platform from strategic partner Sorbotics.ai, drives real machine learning solutions for industry. Their applications are proving to be indispensable tools which solve complex real-world problems and create genuine return for the investing client.
A focus area for the ITG team has been utility operations and building management systems that include industrial refrigerators, chillers, boilers, heaters, fans, compressors, turbines, engines, transport systems, dehydrators, kilns, furnaces, membranes, heat exchangers and many more. These systems are tremendously important for the manufacturing, storage, and transportation of many products but they are notoriously less than efficient even when deployed inside of an environmentally controlled facility. This equipment features control systems but those can be diverse depending on the supplier brand and vintage. ITG is well positioned to work with these systems as their extensive background in controls allows them to apply the machine learning platform without disrupting the native controllers.
If a client’s target process does not have modern controls, or they are otherwise inadequate, then the ITG team can deploy a control solution that encompasses the base equipment while providing a connection point to the SORBA system. “System control must be able to provide base functionality and an interface to our ML platform”, says Ian Weller, Analytics Project Manager for ITG. “Unless there is something non-functional about the boiler or chiller controls then we do not disrupt the existing automation platform. We function as a supplemental, focused piece which can be individually enabled, optimized, and even disabled if needed”, he continues.
One ITG client has realized this benefit in more than fifty sites around the globe. This client, one of the world’s leading brewing companies, has dozens of facilities which feature equipment from both leading manufacturers and local designs so finding a solution that could work across all platforms was a critical design parameter. The ITG team has successfully deployed the platform and their Smart Process Optimization Control Modeling (SPOCM) throughout the client’s ecosystem. “Once the information can be tagged within SORBA, that platform becomes responsible for capturing complete, unbiased training data which feeds directly into the algorithm. The results are accurate predictions of future behavior and optimized parameters which are sent back to the system controller for action”, Weller continues.
“Even though the equipment is similar in product output, there is significant variability in performance based on where in the world the system is deployed and even where inside of a facility the equipment resides”, says Weller. He cautions users that results may be quick, but they are not likely to be instantaneous. “One key concept that buyers must accept is that machine learning requires just that: time for learning”, he states. The algorithms within SORBA are pre-set with information about the target systems, in general, but any machine learning solution requires intense, real information so the results are constantly refined and refocused. Typically, these systems need to see a broad mix of standard and extreme conditions to truly hone the algorithms.
“Depending on exposure to variable environmental characteristics, we’ve seen these systems generate between 5% and 15% increase in overall boiler efficiency. Many sites have seen an overall energy consumption reduction of 5% - 10% in less than 90 days of when the algorithm assumes control”, Weller continues. The SORBA solutions have enabled the development of more complex and accurate models. To date, this client has relied on ITG for solutions at more than twenty sites and the rollout plan includes upwards of hundreds more.
AI will continue to drive the economy and adopters will need an implementation partner with a proven track record. For the client in this example, the work of ITG and the SORBA platform have been an enormous success. Their blend of general automation systems exposure, their deep track record within the ML space, and their proven results make them an ideal partner for someone with a bona fide need for a partner who can deliver.
It should be no surprise that the industrial space offers rich opportunities to apply this technology. Factories vary in the mix of product-specific equipment, but it is difficult to find a facility that doesn’t include systems for managing the environmental properties of areas or processes within. Where consumer applications may rely on inputs from humans and other complex derivatives, modern industrial systems are made of connected components which can provide reliable, consistent information to a modeling system and easily respond based on commands from a learned controller.
ITG Technologies (ITG) has been actively helping clients in this space since 2014. The company has existed for more than twenty years, and they feature a deep and diverse background in the general fields of factory and industrial automation. That experience, when coupled with the SORBA’s IIoT platform from strategic partner Sorbotics.ai, drives real machine learning solutions for industry. Their applications are proving to be indispensable tools which solve complex real-world problems and create genuine return for the investing client.
A focus area for the ITG team has been utility operations and building management systems that include industrial refrigerators, chillers, boilers, heaters, fans, compressors, turbines, engines, transport systems, dehydrators, kilns, furnaces, membranes, heat exchangers and many more. These systems are tremendously important for the manufacturing, storage, and transportation of many products but they are notoriously less than efficient even when deployed inside of an environmentally controlled facility. This equipment features control systems but those can be diverse depending on the supplier brand and vintage. ITG is well positioned to work with these systems as their extensive background in controls allows them to apply the machine learning platform without disrupting the native controllers.
If a client’s target process does not have modern controls, or they are otherwise inadequate, then the ITG team can deploy a control solution that encompasses the base equipment while providing a connection point to the SORBA system. “System control must be able to provide base functionality and an interface to our ML platform”, says Ian Weller, Analytics Project Manager for ITG. “Unless there is something non-functional about the boiler or chiller controls then we do not disrupt the existing automation platform. We function as a supplemental, focused piece which can be individually enabled, optimized, and even disabled if needed”, he continues.
One ITG client has realized this benefit in more than fifty sites around the globe. This client, one of the world’s leading brewing companies, has dozens of facilities which feature equipment from both leading manufacturers and local designs so finding a solution that could work across all platforms was a critical design parameter. The ITG team has successfully deployed the platform and their Smart Process Optimization Control Modeling (SPOCM) throughout the client’s ecosystem. “Once the information can be tagged within SORBA, that platform becomes responsible for capturing complete, unbiased training data which feeds directly into the algorithm. The results are accurate predictions of future behavior and optimized parameters which are sent back to the system controller for action”, Weller continues.
“Even though the equipment is similar in product output, there is significant variability in performance based on where in the world the system is deployed and even where inside of a facility the equipment resides”, says Weller. He cautions users that results may be quick, but they are not likely to be instantaneous. “One key concept that buyers must accept is that machine learning requires just that: time for learning”, he states. The algorithms within SORBA are pre-set with information about the target systems, in general, but any machine learning solution requires intense, real information so the results are constantly refined and refocused. Typically, these systems need to see a broad mix of standard and extreme conditions to truly hone the algorithms.
“Depending on exposure to variable environmental characteristics, we’ve seen these systems generate between 5% and 15% increase in overall boiler efficiency. Many sites have seen an overall energy consumption reduction of 5% - 10% in less than 90 days of when the algorithm assumes control”, Weller continues. The SORBA solutions have enabled the development of more complex and accurate models. To date, this client has relied on ITG for solutions at more than twenty sites and the rollout plan includes upwards of hundreds more.
AI will continue to drive the economy and adopters will need an implementation partner with a proven track record. For the client in this example, the work of ITG and the SORBA platform have been an enormous success. Their blend of general automation systems exposure, their deep track record within the ML space, and their proven results make them an ideal partner for someone with a bona fide need for a partner who can deliver.
Start Modernizing your enterprise today.
