Published on 12/19/2016 | Strategy
In my previous post, I talked about the new approach to technology that we refer to as Big Data/IoT. Let us now look at the few ground level problems created by the previous cycles of OT and IT evolutions and revolutions. At the same time, we shall discusshow theseissues can now be solved and turned into opportunities by these set of new technologies. Needless to say that the use-cases discussed are about various relational and time-series data sets that exist in any manufacturing operation.
One of the most powerful and most widely used tool in the industry is Microsoft Excel. Every manufacturing and business operations produce 50 -100 excel reports per day. These Excel sheets do not only contain data; these are also knowledge repositories.A lot of industry expertise is locked in the formula bars, macros and procedures of the Excel sheets. Most plants usually archive Excel sheets of the last 5 to 10 years. These Excel sheets are then used to make new Excel reports such as “what is my ATF production or sale over last five years? By plant? By customers? Etc.” This process involves tedious manual work which is fraught with errors and delays and also suffers from deliberate fudging of data. Today, one can easily turn these pile of Excel sheets into a data asset that can be “googled” to provide answers with analysis, trends, and sophisticated reports with a click of a mouse button.
Demolition of Data Silos
As I mentioned earlier, there are multiple IT applications and “attached data silos” in any organization. For example, a typical petroleum refinery would have a full suite of ERP applications, Planning, Scheduling and Trading applications, a multitude of MES (Manufacturing Execution System) applications for running the operations such as Laboratory Information management system, Maintenance management system, blend management system, terminal operations management system, process data historians, etc. All these IT applications have attached databases that hold data of the last 5-10 years, depending on the company policy and statutory requirements, after which the data is discarded.
There are a few very interesting changes that are happening to these datasets. Enterprises are waking up to the fact that these datasets are company assets and have the potential of adding major value. They are then jolted by a rude shock that many a times they had signed off ownership of these datasets to the application vendor (legacy licensing agreements!!). It is like signing off the IP rights of your novel to Microsoft as you had written it using MS Word. However ridiculous it may sound, it is a hard reality of today. Enterprises are also realizing that while data in isolation is also valuable, the value increases in geometric proportions when data archives from multiple systems join their forces. While there are many ways of doing it, bringing old archives (of sensor, planning, transaction data) in a common “Data Lake” is getting popular.
Merger of IT & OT (Information Technology and Operations Technology)
Historically, automation systems (PLC, DCS, SCADA, etc.), and Information systems (personal computers, servers, related operating systems and applications) have evolved separately, and till today, mostly use a different set of hardware and software. This happened due to technological limitations of then available (1960s - early 2000s) hardware and software. There was simply no way to cater to both deterministic real-time control requirements and batch data processing requirement of IT on a common platform.
Today’s automation hardware and software do not work across vendors. If you opt for a “xyz” DCS, then you are locked in end-to-end from I/O to Central Controller and HMI in terms of both hardware and software. In many cases, you are locked-in to a particular series of hardware of the vendor and any upgrade may mandate end-to-end change of both hardware and software.
This is gradually changing and PC based PLC’s (e.g. Beckhoff) are getting popular. Since 2005, a new generation DCS from SIEMENS (SPPA-T3000) with Java based design has gradually taken over the market share in power generations and has become the most dominant DCS in that segment. Mega corporations such as ExxonMobil is actively pushing for “Open Automation” for the last 3 years (in this context, please refer to the url - http://www.automationworld.com/exxonmobil-makes-another-push-overhaul-process-automation ). The German government initiative, “Industrie 4.0 (I4.0)”, is also pushing for interoperability so that manufactures are not forced to get locked-in to one automation vendor. They should have the freedom that exists in the IT world (which we have taken for granted and are not always consciously aware of, i.e. do not have to buy PC, printer, router, OS, applications from a single vendor).
In the real world, there is no segregation between IT and OT, especially as far as data is concerned. For example, a typical motor driving a pump will have associated data such as rotational speed feedback (for variable speed drives), winding temperatures, bearing vibration, lubrication oil sample analysis, engineering data, maintenance work orders, spares, financial information (purchase price, book value, etc.). While all these datasets describe one single physical object, they were segregated into automation, MES and ERP data silos. Today, any analytics application that is trying to create a device shadow (aka digital twin) of this equipment or trying to predict a potential failure, has to deal will all these datasets irrespective of the silos they belong to. This is something that is already happening. Going forward, in next few years, hopefully the automation systems will be open and interoperable making life easier while boosting productivity and innovation.
Others
There are various other technologies, such as 3D printing, etc. which are already having impact on performance of manufacturing and production businesses. 3D printing is revolutionizing discrete manufacturing industry. This will also leave a profound mark on hydrocarbon industry. To start with, various physical spare parts will be replaced by STL files, which will be printed on demand by a local 3D printing services provider. Needless to say that this will also affect the logistics industry.
A typical end-to-end scenario may be as this one –
One large turbine manufacturer is already using 3D printed burner heads, and GE recently bought two 3D printing startups* for U$1.4 Billion.
In my next post, I’ll dig deeper in some key use cases for Refiners such as Yield Optimization, and Energy Efficiency.
Thank you for your time.
This article was originally posted on LinkedIn.
One large turbine manufacturer is already using 3D printed burner heads, and GE recently bought two 3D printing startups* for U$1.4 Billion.
In my next post, I’ll dig deeper in some key use cases for Refiners such as Yield Optimization, and Energy Efficiency.
Thank you for your time.
This article was originally posted on LinkedIn.
