JASUBHAI GROUP      ABOUT CHEMTECH     ADVISORY BOARD     AWARDS       EVENTS     PUBLICATIONS     CONTACTUS    
Chemical & Processing
EPC
Oil & Gas
Refining
Automation
Pharma Biotech
Shipping
Power
Water
Infrastructure & Design

Cloud Computing and Plant Performance
Animesh J Mathur, General Manager - Marketing DCM DataSystems. Advancements in IT have brought radical shift in data transmission from conventional wired networks to wireless transmissions. Huge bandwidth capacity, massive microprocessor powers and 4G networks soon to be reality, the author shares his perspective on how cloud computing will bring about mammoth change in data processing in chemical plants and the pros & cons associated with these services.

There was a time when Industrial Automation technologies - especially on the communication front - were much more reliable than the networking technologies which were available in the IT field. RS 485/RS 422 Modbus etc. used to work quite reliably even in ┬noisy╩ industrial environments without a loss of integrity of data. However over the last 10 years the massive growth of the Internet and advancements in telecommunications put together have created a totally different paradigm.

To get a perspective of the mammoth scale of change that is in the offing it will be good to look at the analogy of the Power Generation Utility companies. The power generation scenario in the mid-1800s was dismal and Edison had not yet invented the incandescent bulb. If someone had to use electric machines, they had to generate their own electricity. But by the early 1900s, utility companies like the Chicago Edison had started generating enough electricity which could be transmitted and supplied to factories and homes over long distances. Who would have thought of washing machines in homes and electric trains a mere 100 years back?

A similar change is getting unleashed at us by the creation of huge bandwidth (capacity) telecom networks and the massive power of today╩s microprocessor. Due to the telecom networks today you can read your mail anywhere and update your status on a social media page on the move.

Presently most of the discussion around ┬cloud computing╩ happens around the IT landscape of the company with each IT company espousing their own thought process. At the base of this whole model of cloud computing are the following key things:

1. Shared resources;
2. Large capacity of transmission;
3. Utility based operation where you are charged for usage instead of spending on Capital Expenditure.

If we were to look at any chemical company you could break it down into two parts - one is the Plant and Plant based or supporting facilities and the other is the commercial activity of marketing, billing, distribution, procurement etc.

There is enough and more being written on the usage of the cloud in the second area so we will focus on what possibilities and challenges exist in the adoption of the cloud for the ┬Plant & Plant based or supporting facilities╩ Within the Plant and the Plant Based operation we may be able to break down activities into broad areas such as Project Design/ Product Design/ R&D/Project Management; Production Planning and Execution; Storage of intermediate and final products and by-products; Factory Floor and Field Level Automation; Maintenance of plant and equipment.

The First 2 areas already have a considerable availability of software on the cloud and if someone does not want to invest in having their own systems they could search and get pay-per-use applications. This is a constantly evolving field and we should see more and more applications here. On the other hand SCADA (supervisory control and data acquisition) and PLCs (programmable logic controllers) are still something, which are a distance away from getting the benefits of the cloud.

With 4G networks to become a reality soon across India you will have low cost broadband capability. With field level transmitters having enough intelligence embedded in them, it is a matter of time that they will start transmitting their information via wireless networks or public networks. Similarly there will be actuators, which will be able to receive information via wireless or public networks. So, you could have a complete feedback loop working using public setups. And thus the whole concept of wiring will disappear.

If that happens then it will technically be irrelevant where the Programmable Logic Controller (PLC) is physically located. You could have a massively powerful PLC being shared between multiple plants and all computing being done at a central location. You could then actually have providers who run and manage the PLCs without small plants having to invest in the infrastructure of doing wiring and programming PLCs etc. This is in principle same as having a third party email provider like ┬Gmail╩ providing email services. Now the small plants will not need to invest capital in instrumentation equipment and working capital in instrumentation people.

The Programmable Logic Controller (PLC) service provider would come and install the transmitters and actuators and you could just give them the various recipes that you intend to run. Since the service provider will work with so many companies he will appropriate the costs of the capital equipment across various customers. Due to this two things will happen ?? one will be the lowering of costs and second companies will pay only on a usage basis so the smaller companies will not end up with large depreciation costs on their books.

From a maintenance perspective, it will make the life so much simpler - since transmitters and actuators can already communicate via public networks, the makers of these transmitters may have a central monitoring system, which can keep monitoring the health of these devices, and ┬on the fly╩ upgrade their internal software. Even if we don╩t reach that stage there are SQC and maintenance packages available where you can automate the procedures and they will alert you to do planned shut downs.

While this Utopia like situation can happen in the very near future, there a whole lot of roadblocks enroute. First would be the standardisation of protocols - all the transmitter and actuator companies will need to use standard protocols for transmission of data on the cloud. Second would be the security - security to ensure that the service provider does not endup using proprietary data. Security is the biggest concern area even in the large scale adoption of the cloud in the IT side of things and the government of India has setup a task force to suggest the standards and laws for the ┬cloud╩. Third would be the human aspect - the adoption of the cloud would displace some kind of existing job roles and create totally new job roles.

While large chemical companies have outsourced production to smaller companies the cloud will bring about an outsourcing of roles. And finally the biggest roadblocks will be in our minds. We will need to start by testing the cloudbased services on peripheral areas and then as we gain confidence we can move to the core.

A lot of the journey to the cloud is right now unknown but the human mind has the capability to figure out solutions and progress forward. It will be the same with the cloud. Wish you a safe flight!