How an Industry Veteran is Revolutionising Manufacturing with Data Science

Interviewer:

Good Evening Mr. Subramaniam Mohan and welcome to this series of best BDM capstone projects. Firstly, Congratulations on being nominated for the best capstone project. I would want you to introduce yourself to our readers so that they get to know about you.

Subramaniam Mohan:

Good Evening!!! My name is Subramaniam Mohan, I graduated in 1978. I believe many of you were not even born then. I am from PSG College of Technology Coimbatore, affiliated to Madras University. My basic graduation is in mechanical engineering. Subsequently, I qualified in accounting, AICWA and then Graduation in industrial engineering from IIIE-NITIE, Mumbai. Speaking of my professional experience, I started as a project/maintenance engineer at ACC Madukkari, Coimbatore. Then, I gradually slipped through into engineering, then projects, mostly spent in Oil & Gas and LNG industries. I have 41 years of Industrial experience. I worked 15 years in India and 26 years abroad. In India, I worked for ACC a cement company. Then I went into the steel industry and joined Rourkela Steel plant (RSP) which is the first integrated steel plant in the public sector in India. My next stint was with Dunlop – a tyre factory. Then I moved into the Fertilizer plant Southern Petrochemical Industries Corporation (SPIC) Tuticorin/ Chennai. These are my 16 years of experience.?

You would be wondering how I just moved through different industries segments. When I migrated to the Middle East, I started with Abu Dhabi, then Dubai and then moved to Muscat. I spent quite a long time in Qatar as well. From Qatar, I moved to Japan. From Japan as a base, I started Shuttling between France and the US and Doha, Qatar. That was around 11 years, it was a continuous project experience. I must admit that it was a fantastic experience gained through working with different cultures. The Japanese are workaholics, they're really spend 16 hours a day. The French really want to work-rule guys. They used to work around 30 hours to 35 hours, not even 40 hours and the US is in between, And we are a mix of everything. That's how our projects were going on. And managing those people itself is a big challenge and it was an interesting experience. And we continued to be awarded with the projects and we were happy to manage those resources.??

Interviewer:

Thank you for that quick introduction, I must admit that you carry a wealth of experience. Moving to my next question, how did you learn about the Data Science program and what was the inspiration for you to enroll into this program?

Subramaniam Mohan:

After coming back to India, I thought of really starting my passion. As an organic farmer, it's new for me to really move from One established, comfortable field into a totally new field in the unorganized sector. So, though, I come from an agricultural background family, but I never tried agriculture in my life, until I started this time. So it is a first time attempt/ venture. I went to Tamil Nadu Agriculture University, where I took a short term course to understand and get networked with people like me. I was surprised to see hundreds of experienced engineers migrating to farming. Who really came down from industry to or create their own farms. So that's where I met a lot of people. A lot of farms. and there was Isha Center for Agriculture promotion doing great services to the budding farmers. Isha conducted yoga classes in a residential agricultural program in various established farms that gave me a lot of Exposure and contacts with the farming community. An opportunity to meet a lot of retired farmers from various parts of India.?

Then I could do that for four years and that plot of land was taken by the government for airport expansion. So I got to give away the land. So when we were left over the small portion, we couldn’t really continue, then I started searching for alternative passionate engagement. I just didn't know what to do. There are about 20 small scale industries here and I could ? ? give them some management/ productivity enhancement advice as this was something that I was fascinated with. It is around this time when browsing through the internet I stumbled upon IITM’s BS Data Science Program.

So I have a lot of passion for data because my work demands it. Most of my work is - ?whether it is a design, engineering or it is a product engineering or plant engineering, data everywhere. In plant engineering, we have to deal with AutoCAD, and in Product Engineering I have to manage business with Intergraph Microstation data. We go with the 3D model for everything. And even for fixing the maintenance problems, we really go with the vibration monitoring data and pipeline vibration, piping vibration data.

So eventually we end up with the data everywhere. But all my analysis, everything was done on Excel as a tool and very little opportunities for multiple tools management. In addition our business started moving into SAP compelling us to really pull the data from SAP. Eventually the front end always excels. Later on only, we slightly started working on tableau and Business Information Systems in SAP. When I saw this data science program and I saw the curriculum materials, I really thought that it is a course for me, I just immediately jumped over there.

Interviewer:

Moving to our next question, how has your journey been till this point in the program.

Subramaniam Mohan:

When I started with all this academic background and experience, I reckoned it was a simple cake-walk. But it was not like that. I could not count it for supporting academic pursuit ? ? . It's almost restarting, learning after a long time and starting on the Books. So when I saw the curriculum, I thought that it would be easy for me. So I took four subjects in a term. Then after writing only I realized that it is not a breadth oriented curriculum but it is a depth-oriented curriculum.

So I realized it is a very wrong plan to really rush all the things. That it should be a really a phased manner of progressing, through the courses. Then I changed my plan. I moved over to the three subjects a term. Which really moves around the four to five hours a day means around 30 hours a week. That was a comfortable thing. What I expected for the three subject program it took a while for me to really clear the foundation because of my wrong plans. And that's where most of the students will start with a lot of expectation. and with the desire to finish it as quickly as possible. That's all, they will get stuck with these overloaded programs. ?????

As I said that, I have a lot of liking for data. Also, because of my long working and Moving around various parts of the world, I met a lot of people and I have a lot of data in hand. Which basically only I can really decipher and put it in any use. So now I need some tool for really stacking up the data in the most reliable way. And I need to have smart tools and how to really convert the data into visual models so that the data can make sense through applying ML modules like MLF, MLT, MLP? It's handy too. It is the tool which we need to have, and this subject knowledge and the knowledge on the tools. It is a very essential thing to really use and exploit the data and the programming on Java and Python are an essential language to really build the tools.

Interestingly, I do have a lot of Involvement in Share markets. I started really picking the shares based on intuitions. I didn't really do any structured analysis of the data and I never really read up on the fundamentals. But I wanted to really try it on my own data analysis, which I really collected over the period of around 20, plus years of experience. This expectation? really gave me a ground towards pursuing Data Science. This would help me to do my investment robustly. So, these are all the reasons behind why I went on the IIT Data Science Program.

There is another mistake that I have made is I thought of really pursuing both the diploma, most people do that. So I mixed up the subjects. both from Data Science and the Program. I took a couple of exams for about a year. I found it and many of my friends and colleagues were then saying they also tried this model so that they can balance it evenly. I felt that the subjects of data science is a little easy. I am a non-programming background guy. Though my work had heavy interaction with the IT department, I was always a business side guy; never went out to the coding side. So even now I find coding to be a difficult domain. Then I listened to my colleagues' suggestion that let's finish one at a time. By concentrating on data science first and then on the? programming side. So, I strongly advise the students not to mix both domains. My suggestion is prioritize between these two, pick one at a time and finish it. Then move on to the second one. It matters for the different people because people with a programming background could be doing differently. So even before I was starting on the IITMs program, I started on NPTEL, I started using a personal email ID, because we got the IITM email only after we finished the foundation. But NPTEL as welI pursued it without the knowledge of IITMs Program. So, I didn't change my email until I finished my foundation. By the time I finished most of? the courses towards the Data Science Domain Certificate. I started really giving exams and finished, I think out of six subjects I finished about 5 subjects required to really qualify for the domain certification. Now, only one more subject is pending. But in the meantime, when I started the IITM program the course materials demanded more time. So, I deferred the NPTEL program for some time until I really finish at least my two diplomas from IITM.?

So, this is a background about how. I really moved into this Right now, I am 68 years old I have no concrete idea about how I'm going to use my late acquired knowledge but I'm sure that, any knowledge is definitely of use only if you really find out the right opportunity, And thankfully, that BDM project really compelled me to go on the line with one of the existing industries which had a problem written and the owners were struggling hard to read, find out a solution how to really manage it over for the rest of the time.?

Interviewer:

Thank you for sharing your experience of the program. Moving on, to our next question, you did mention that the BDM project was around a problem that was very evident but the owners were not able to identify and find a solution. It would be great to get some more insights on this, your approach, methodology, challenges if any that you faced during the course of the project

Subramaniam Mohan:

So this is exactly where I went into one of the manufacturing industries. The name of the company is Prema Precision Tools. And they are basically job order processing guys and they have their own six, CNC machines (Computerized numerically, controlled machines). They have around 6 to 8 machines, around 20 to 22 people. And they were operating most of the time 6 days a week, 24 hours and taking one day for shutdown in a month. The workers are involved in shutdown activity, So they are not given any vacation or leave. So the factory was doing extremely very well. It was promoted by a first generation entrepreneur. He is also a seasoned guy and he came back with a lot of expectations with the Make in India campaign and there was Entrepreneurship Development program and the startup aid from the government around 25% subjected to a maximum of 25 lakhs of foreign investment minimum of 1 Crore rupees. So he qualified for all those things, when he put up this factory and the factory is 1 crore plus investment. The working capital requirements for three months at least requires around 30 to 40 lakhs. As I said, he is a technical entrepreneur. He is not a seasoned industrial promoter, so he was expectedly a technical solo who could focus more on the Technology behind the conversion part. He was under the impression that once he managed the technology he was successful in the business. But later on after almost six months, he started realizing that the technology part is only one of the many facets, which need to be managed. Then, that is the time where we both met and we started discussing it. I was explaining that I do have multifaceted qualification experience. I could see that there were a lot of opportunities to really modify your work and we can improve it quickly within three to six months. We could see a turnaround when we started working.

But when I went into the industry, the workers started thinking that one of my exercises is going to end up in reducing the number of people. Anyway, any external exercise, whatever comes inside, the factory people, the workers generally skeptical think that this is a cost reduction, eventually it ends up in elimination of certain people. Then I took a lot of effort in explaining to them that my Work is going to bring in a discipline to simplify the work and improve productivity. ? ? ?It will be the same work that they are doing But it will have a different approach so they don’t need to worry about it and my work is to really help them with how to really modify the work culture. In a way to increase the Productivity, it took a while.

I had to ask the Promoter to allow me to really put up some activities that promote the team. management, and the operators should feel that they complemented each other and not really going for any punitive measures. So almost every weekend we had dinner together around eight o'clock with the team from both shifts because this company operates and a 12 hour shift basis. So, it was the transition time of the shift. Employees from both shifts will be there.. We had dinner and that was the need of the moment where the fellow workers were really happy to be with the management and they started opening their hearts at discussing freely. This was very helpful in data collection, and subsequently, completed whatever the request, we made them to ?????prepare.

Then the logbook writing habit is improved. And that is where we really pulled out all the data. Then we started showing them the details around the material coming in and the finished product that is expected and the steps in the operations that can be improved which are important to improve productivity. So, when you put it up in the paper and show it at a higher level, then people appreciate it very well and start a different work method.

There was scrap generation which was quite higher than normal. Which eventually ends up in consuming more tools, more consumables, and more power also. So, which is unwanted for anybody and it doesn't really bring in any benefit to either management or to the workers. The turnaround cycle was also very big, we addressed all these things and then we gave a modified workflow which helped reduce the turnaround cycle time.?

The subcontractor and then prime contractor work was not seamlessly tied up together. There were frequent material handling, that doesn't really add any value to the job to be completed. So we modified how the transport schedules really are connected and how the materials should move from one shop to the other shop. Even from the main supplier's shop itself, we started requesting the main supplier to streamline the job flow. We requested them to bring it to our Subcontractors facility and we would take the documentation of material receipt from their side, from the subcontractor. Then we will start working and getting the material. Then we will finish our part of the work and intimate the client. The normal workflow was pulling the job from the main clients shop to our shop from our shop, then sending it to Subcontractor from Subcontractor again, getting it to our shop and then sending it back to the main product manufacturers’ Shop. So, this unwanted material moment has saved nearly around 20% of time, the transport cost is less. And this is what this is the biggest thing we have done in the BDM project.?

Coming to the data collection and methodology for the BDM project, we pulled out the data, cleansed the data then normalize the data. Because I was not good at Python or Java around that time since I had not finished my Machine learning courses I did most of the analysis on Excel. Though I have previously worked on SQL and Access Database I have lost touch now as it has been a while now that I worked on these tools. In addition to this, there was no ? ? ?fiancial comfort of ?????putting up a ? ? ?SQL database because the owner of the company always gives more priority for any jigs ? ? ?and fixtures ? ? ?, ? ? machines or tools which he feels are for building a world class process. But anything on the computer and ?????data related things, he always feels that it is a waste exercise, it doesn't really produce anything tangible. So he was always inclined ?????to differ from all these data management tools and I couldn’t really handle both together - ??????the disciplining of the shop floor and disciplining of the office systems. So between these two, I felt shop floor modification is easier for me to handle because the technical promoter is more inclined to accept that one modification. Anything I really gave to office systems will not fly easily. So what I felt is let us leverage the most of the tally. Tally is the accounting tool. He was using it. So tally do have a lot of reporting mechanisms where I explored the Tally mechanisms and pulled out the data and cleaned and started using Excel. So finally, all that information is really converted into the visual thing either in the pie chart or Parreto charts or in the histograms and all this things are easily showing the power of Data management. That meaning of the data behavior? ? ? ? ? ?on performance ? ? . So that was really accepted. Even the normal operators have understood and started really appreciating it very well so that has really given already acceptability to move away differently from what they were pursuing. That's about the shop floor at the BDM project.

Interviewer:

That’s been very insightful. Thank you for detailing out your project so beautifully and patiently. Coming to my next question, given the wealth of experience you carry and being a part of this program this far, what is the one message that you want to give to the students of the program and to our readers?

Subramaniam Mohan:

Okay. now, having come this long I will have a lot of messages to share with the students And my key message is Data science is that subject, which is definitely needed, no matter which domain you are going, whether it is medicine or trading or logistics or production, or products, or contract management or product management or procurement management. Doesn't really matter, you always generate Plenty of data and your knowledge is always built based on what you really win the world. When you look back, you really appreciate it, What you have done, and what you could have done better than what you did. So for that, you got to really have a good data analysis. So the curriculum, what BS Data Science is, is very much aligned in that process.

Yes, all subjects are not equally loaded. Subjects are loaded in varying degrees, for different background of the people. If people who come from a descriptive experience. They choose different subjects easily and the people who come from a programming basis, they choose it differently. But it has a mixed bag of both. So you accept it is taking you in the right car and you have a great future.

Having said that, I'll give you a little bit of specific tips. Foundation is a child's play. You really enjoy doing it and you have to spend not much time and you can spend the minimum time. and of course, math has its own difficult pitch, so you have to give a little bit of extra effort, a little bit of more importance. This is the subject the way it is. But when you come to the diploma, As I said early. Don't take it all at once, foundation 4 subjects is not at all a problem for subjects. But this term is not a problem. You can definitely take it. and in order to excel you need to commit 30 to 40 hours a week, No sacrifices on that. And if you have the capacity of really giving that one then, go for it and if you are really having, Committed. and you have time left out, then scale down your number of subjects information. Okay, even if you fail to pass one or two subjects, don't worry about it. Go for it again. You will definitely make it. and diploma. On the other hand. Is not easy like foundation but it is more interesting than foundation. so, you decide which you want to go into, whether it is in the data science part, or in programming. My strong suggestion is take two subjects at a time.

Any subject with the project? I always felt that the project moves faster than the theory. Means you really go fast. Without the theoretical knowledge and end up doing the project, sometime after two or three weeks your project will be advanced. Then the theory is part of what is required to redo. The project is not going to help with any theoretical knowledge. So, my suggestion is always, Finish the theory in totality and take the full grasp of what you really had. made it on the theory.

Then those who do the project and the following substance immediately. In the following semester, you do that project work. You really enjoy it and finish it. Oh yeah, when you are doing the project, the very important thing is the rubrics. You got to really adapt to the rubrics, rubrics is the gospel, there, and we got to adhere and deliver the deliverables. Strictly in adherence with the road race. If you have the best capacity, go beyond, then you will be accepted. and then the professors, he is the inspector. All the people are really friendly and they provide support. Whatever you guys need, In addition, you have your own friends. There are a lot of user groups active and a lot of Good new ideas. So pick that one and start working on it. You will enjoy it. And don't miss the paradox annually.

Even paradox, where you will meet all these guys, so that gives a different togetherness. It also really makes you really connected to the people and discusses the subject ways and really finishes up. These are my few suggestions for you guys. If you really want to make the program interesting, finish it successfully. It is what I am suggesting, I'm not pushing that. When you always are successful, there are a couple of slips that will come in between, don't be certain. Take it challenging and find out what is that Went wrong in your earlier submissions. and put it back and deliver it. It will always go but give importance, always to the project, project is the real thing for you to really test your own subject, knowledge and how you are really applying it in the industry. How well you are doing the project really shows how well you are going to perform in the real time jobs. All the best guys!!!?

Interviewer:

Thank you Mr. Subramaniam, for your time, for sharing all your experience and views with our readers, It’s been a pleasure hosting you. Have a great day ahead.??

Subramaniam Mohan:

It’s my sincere happiness, I want to really show it to you. I really enjoyed it very well. And anytime I am ready to share my Little bit of knowledge, whatever has been acquired. through this curriculum and as well as my previous experience about those I'm always available to anyone.?

I will take this opportunity to express my gratitude to Dr. Andrew for providing me this platform to share my experience. Ms.Bharati who has been driving this entire program. Prof. Ashwin and Dr. Aaditya and the entire TAs for all the support and guidance throughout the BDM Capstone Project journey.??

Thank you and have a great day.