Representative End User Clients
Representative Automation Clients
Representative Software Clients
Last week a significant acquisition in the industrial software markets took place for process manufacturers Oil and Gas and Chemicals producers. AspenTech, a market leader of Engineering Simulation, Manufacturing, and Supply Chain software, made a very strategic move by acquiring Asset Optimization software supplier Mtell. However, the uniting of these two companies begs the question: why would a company whose primary business is engineering simulation - with the majority of its customers in engineering design and process optimization - be interested in a company that serves the asset optimization and maintenance markets?
Mtell's provides solutions that help reduce operational risk, end breakdowns, and increase net process output and profitability by using machine learning and artificial intelligence, while AspenTech's solution expertise is in engineering simulation with strong connections to Academia. Mtell's software and unique approach to making machines smarter by predicting asset failures with a bit more sophistication than traditional condition based monitoring. Alternatively, AspenTech has continued to improve process simulation technologies, improve design efficiency and decrease manual effort by engineering disciplines.
There are likely several reasons for the acquisition, but the short answer - this is all about using "Big Data" to help operations processes become more predictive in nature. The use of Analytics in particular, is helpful in bringing together the process engineers who understand industrial processes with the equipment and asset engineers who understand how assets and machines work. By creating new platforms, introducing machine learning algorithms, and operational analytics software, experts from both disciplines may become more proactive in operational and maintenance practices.
Many studies have shown that more than 80 percent of asset failures are random in nature. Boeing has completed studies that reveal 85 percent of asset failures occur in spite of calendar maintenance, while Emerson process believes that 63 percent of all maintenance is unnecessary. Many manufacturers have only been able to make marginal efforts in predictive or prescriptive maintenance practices. ARC studies and surveys show that less than 3 percent of maintenance practices are proactive. Why is this the case? This is because maintenance experts have had a limited view and skills to understand and interpret the process data that ultimately affects the reliability and longevity of the assets. In other words, assets may fail because of the way they are operated and experts have not been able to detect this in their data. The adoption of big data, analytics, and machine learning is relatively new, but industrial organizations are quickly getting up to speed on this intersection of data science and industrial knowledge of both process and assets.
At the AspenTech Optimize software conference back in 2015 the company announced plans to invest in process-specific analytics for the future of the aspenOne solutions. The company spoke about how equipment analytics could help optimize assets by allowing users to predict unexpected outcomes and provide prescriptive guidance for operators or engineers. They described the potential addition of how a non-historian-centric database might provide context for both structured and unstructured data. This new “big data” element would support prescriptive and predictive analytics in conjunction with the time series data. For industries served by AspenTech, process analytics would allow users to examine process flow data from a historian with data from engineering such as that found in the plant schedule or from a piece of equipment in a HYSYS flow sheet. In the process simulation ecosystem, this would bring with it a multitude of opportunities to build out data sets outside the realm of normal process data stored.
For AspenTech, the acquisition of Mtell is one step closer to making prescriptive operational guidance and prediction of unexpected outcomes a reality.
What I have learned about analytics, machine learning, and prescriptive operational guidance use cases so far:
Mtell's provides solutions that help reduce operational risk, end breakdowns, and increase net process output and profitability by using machine learning and artificial intelligence, while AspenTech's solution expertise is in engineering simulation with strong connections to Academia. Mtell's software and unique approach to making machines smarter by predicting asset failures with a bit more sophistication than traditional condition based monitoring. Alternatively, AspenTech has continued to improve process simulation technologies, improve design efficiency and decrease manual effort by engineering disciplines.
There are likely several reasons for the acquisition, but the short answer - this is all about using "Big Data" to help operations processes become more predictive in nature. The use of Analytics in particular, is helpful in bringing together the process engineers who understand industrial processes with the equipment and asset engineers who understand how assets and machines work. By creating new platforms, introducing machine learning algorithms, and operational analytics software, experts from both disciplines may become more proactive in operational and maintenance practices.
Many studies have shown that more than 80 percent of asset failures are random in nature. Boeing has completed studies that reveal 85 percent of asset failures occur in spite of calendar maintenance, while Emerson process believes that 63 percent of all maintenance is unnecessary. Many manufacturers have only been able to make marginal efforts in predictive or prescriptive maintenance practices. ARC studies and surveys show that less than 3 percent of maintenance practices are proactive. Why is this the case? This is because maintenance experts have had a limited view and skills to understand and interpret the process data that ultimately affects the reliability and longevity of the assets. In other words, assets may fail because of the way they are operated and experts have not been able to detect this in their data. The adoption of big data, analytics, and machine learning is relatively new, but industrial organizations are quickly getting up to speed on this intersection of data science and industrial knowledge of both process and assets.
At the AspenTech Optimize software conference back in 2015 the company announced plans to invest in process-specific analytics for the future of the aspenOne solutions. The company spoke about how equipment analytics could help optimize assets by allowing users to predict unexpected outcomes and provide prescriptive guidance for operators or engineers. They described the potential addition of how a non-historian-centric database might provide context for both structured and unstructured data. This new “big data” element would support prescriptive and predictive analytics in conjunction with the time series data. For industries served by AspenTech, process analytics would allow users to examine process flow data from a historian with data from engineering such as that found in the plant schedule or from a piece of equipment in a HYSYS flow sheet. In the process simulation ecosystem, this would bring with it a multitude of opportunities to build out data sets outside the realm of normal process data stored.
For AspenTech, the acquisition of Mtell is one step closer to making prescriptive operational guidance and prediction of unexpected outcomes a reality.
What I have learned about analytics, machine learning, and prescriptive operational guidance use cases so far:
- Accurate predictions about failures in assets or process upsets require data in context. This is data coming from industrial assets and data combined with the process data upstream of the asset and data from the downstream processes.
- This data in context often requires interpretation by a subject matter expert with knowledge of the continuous process and physical asset.
- AspenTech, engineering simulation models, already contain significant IP and can provide context to build the predictive and prescriptive analytical models.
- Mtell Previse with aspenOne engineering simulation together may provide a new dimension to asset optimization by supporting predictive operations and an entirely new set of work processes that extend beyond the maintenance and reliability engineer to the process engineer.