Amid increasing demand for sustainability and cost-effectiveness, leveraging data analytics and machine learning (ML) has emerged as a transformative approach to predictive maintenance and process optimization. These technologies not only minimize downtime but also enhance the efficiency, safety, and profitability of chemical plants.

In this regard, the leading experts recently spoke at a panel discussion on ‘Harnessing Data Analytics & ML for Predictive Maintenance & Process Optimization in Chemical Plants’ at the CIO Connect 2024 organized by Indian Chemical News (ICN) on November 29, 2024 in Mumbai. The discussion was moderated by Pravin Prashant, Executive Editor, ICN.

Themed ‘Connected Chemical Plant and Industry 4.0’ the CIO Connect 2024 was held in association with AVEVA and Grassdoor.

Speaking on strategic potential of manufacturing data, Namrita Mahindro, Chief Digital Officer, Aditya Birla Chemicals opined, “To improve the reliability of plant and equipment, there is a fair amount of work that we have done at the Aditya Birla Chemicals and other group companies. In terms of looking at increasing reliability, data has been the key. The platforms such as AVEVA and others provide the historian that you have for actually collecting the data and then creating a layer on top of it. That is where you start building your models and building your reliability by predicting the propensity for equipment to to break down. A lot of our businesses have equipment which are legacy equipment and you may not necessarily have OEM support. The second area which becomes extremely important from an operational excellence perspective is quality. That can be achieved with the ability to trace back using solutions, like AI to the source of the problem. The third area is around productivity or efficiency and there again it becomes important to figure out the performance by collecting the data to build those models. The efficiency and cost reduction have seen a significant improvement.”

Speaking on overcoming barriers to OT data integration and driving relative maintenance at scale, Ajay Bageshwar, Head - IT, Godavari Biorefineries said, “There are a lot of barriers in this journey. First and foremost, the trigger point to start the industry 4.0 journey must be known. First of all there should be a handshake between CIO and MD of the company and also the company should be aware about requirements for industry 4.0, the implementation plan and funding. The next point is adoption of the change as whatever change we bring to the organization is inversely proportional to the acceptance. The visualization and the journey must get adopted into the initiatives. From the operator to business level, IT or OT teams, this is very important. This is the first challenge and the first barrier everybody faces but when it starts from the top, things become easier. As we get more into implementation part, and start our journey, we go for digitization and automation in the factory. Third and most important point is choosing a partner that matches your business process, your setup, technical setup and your goal.  In terms of a few technical aspects, data is the most critical point. The second is the communication medium i.e, how data is going to flow from right from the physical layer to the application layer. Third and foremost is security.”

Sachin Khamkar, Senior IT Manager - India, Clariant, “Predictive maintenance is not only important for a company but for people as well. When any of my plant operators, including plant managers are going on a long holiday, they expect a peaceful time, a process where they can sit at home and look at all the data might not look feasible at this stage but we can work towards that. Instead of looking at just what is happening right now, we really need to look for a system that can help us to identify the problem well, before it comes. The mentality has to come up to that level of leadership interest to talk about how we can bring this solution on to the table that can overall help. There are billions of data points which can be used and a platform like AVEVA can create so many communities where the data can be pulled and can we utilized to minimize time spent on redoing the same stuff or reinventing the same stuff to identify how exactly the pattern works. The predictive maintenance is based on the potential of one’s instrument. You can choose the kind of maintenance, whether it could be reactive, proactive or maybe predictive. When we talk about digitization, it is not going to do the magic unless and until done properly.”

Sharing his thoughts on role of AI,ML and data analytics in enhancing efficiency and safety, Krishnakumar Pandey, Head - IT, Epsilon Carbon said: “With AI and ML, we have a model which can take the data real time, analyze the data real time to take real time decisions which will help to into aspects. Firstly, it is going to improve the production and secondly, it can help us to avoid some of the safety incidents. This is a great opportunity for all of us. However, as we speak, there are a lot of challenges when we talk about the data. There are three major challenges in data management. The data generated is quite fragmented and difficult to collect the data all together. This is an important aspect one has to address when starting the journey towards the industry. 4.0. The second challenge is in terms of data quality because AI and ML models need a good amount of data and some of the data is not available. AI models cannot understand or predict properly. This is another challenge where we need to work on improving the quality of data. Then in terms of contextualization missing in data which is required for AIML model to understand and help improve the efficiency.” 

Sharing her thoughts on the importance of cyber security, Ashima Bajaj Seth, Chief Digital and Information Officer, Godrej Agrovet said, “First step in ensuring cyber security is to identify the risks that you are dealing with. Through the risk assessment in the manufacturing setup, we try to address unauthorized access, data breach or disruption of production. But apart from that, there could be other associated risks depending on the nature of business and the manufacturing setup and depending on the legacy that you have. Over the years, we have developed protocol, and frameworks which help us define how we want to address and protect against these risks. The People Process and Technology are important. People because it is the understanding and awareness that we need to be careful about the data that we are dealing with. As an organization and as an individual, it is each and everyone's responsible for security. That is on the culture side of people. For management and people's awareness of the processes as an organization, one needs to have policies and procedures in place.”

Rajeev Raval, VP and Head of Digital Productivity, SML said, “As an Agro input company, we have implemented digital transformation two years back. Given the legacy of 50 years, we had to bring the visibility and convince the top management because alignment is necessary for any digital transformation journey. That was the first step and then we had to identify any low hanging fruit. Firstly, we had to find ways to unify this data and then bring that visibility of this data to various departments.  It was followed by providing the data and its visibility among different departments like supply chain and production in back end as well as front end. With visibility, they could identify their route and based on this, come up with a solution. That was first kind of advantage to convince our management that digital transformation is very much required. To leverage technology in our day to day operations where there is a lot of man power involved, there are certain processes through which we can automate and generate more productivity improved efficiency and reduction in wastage. With readymade data available, we can leverage that in order to generate a more profit for the company directly or indirect way.”