Data analytics researcher María Isabel Meza Silva is helping Irish manufacturing businesses step into the era of industry 4.0, and hopes to see the industry in Colombia follow suit.

Originally from Colombia, María Isabel Meza Silva holds a bachelor’s degree in industrial engineering from University of Los Andes. In 2017, she came to Ireland and joined Irish Manufacturing Research (IMR) as a data analytics intern.

IMR is an Enterprise Ireland and IDA-supported research and technology centre with labs and industrial pilot lines in Dublin and Mullingar. Following a few months back at University of Los Andes, Meza Silva returned to Ireland in 2018 and has been working as a fully fledged data analytics researcher at IMR since then.

Her main role is to help small and medium manufacturing businesses discover the value in their data. She does this by identifying key metrics and transforming raw data into meaningful and actionable insights.

Meza Silva also leads IMR’s outreach programmes with primary and secondary schools, immersing young students in new technologies such as the internet of things, AI, augmented reality and collaborative robots.

‘A common misconception is that analytics only works in large organisations with large resources’
– MARÍA ISABEL MEZA SILVA

What inspired you to become a researcher?

To be honest, it was not part of my plan to become a researcher when I was in university and I knew very little about what working as a researcher entailed. All I knew back then is that I wanted to work in manufacturing and have an impact in society. Since my dad brought me to the place where my favourite cereal was produced, I have always enjoyed going to factories and figured out how things are made and how machines work.

When I joined IMR as an intern, I realised the importance of applied research. IMR’s research is focused on using new technologies to help small and medium Irish companies improve their industrial processes. And if we can make these processes better for companies, why wouldn’t we? At the end of the day, manufacturing companies are the ones that provide us with the food we eat, the phones we use or the clothes we wear.

What research are you currently working on?

Some of the research projects we are currently working on were born as part of the Industrial Data Analytics Framework (IDAF) research project. During this project we realised that the main barriers to the adoption of analytics are generally managerial in nature and they probably stem from the lack of understanding around where to start and how to manage an analytics project, which also translates into a lack of understanding on how to go about solving problems using data science and inherent analytics capabilities.

In order to tackle this, the analytics team developed an industrial data analytics framework that aims to abstract the process that IMR has done with many companies, to produce a knowledge base that could be used as a self-service or guided tool for companies. This would provide an assessment of current analytics capability; a review of projects and techniques that are relevant for the present capability, infrastructure or culture; and the incremental steps that could be taken to bring value through new data analysis capabilities.

Based on the assessment, we have worked with companies to implement analytics initiatives that can bring value according to their current capabilities, and have established steps to increase their analytics capabilities. Understanding the current capabilities and process is key for companies to unlock the value in their data.

In your opinion, why is your research important?

As I mentioned, manufacturing companies are essential for the functioning of society. Think about the impact they have had during the Covid-19 pandemic. Several manufacturing companies repurposed their lines to start producing personal protective equipment while pharmaceutical companies are working non-stop to supply vaccines.

However, some of the processes that are used in manufacturing have not changed in decades because there has not been any need, and the fourth industrial revolution has brought advances that industry cannot ignore.

IMR aims to demystify, de-risk and deliver these emerging technologies that represent an opportunity for companies to make their industrial processes work better. Within the analytics team we work closely with industry to show them how data can be best used to increase revenue, customer satisfaction and product quality, and to make informed decisions.

What commercial applications do you foresee for your research?

There are three areas that companies have shown particular interest in when it comes to using data: predictive maintenance, defect analysis and overall equipment efficiency.

Predictive maintenance allows companies to decrease costs by foreseeing when a machine will fail and take actions before it happens, while defect analysis is being used to increase product quality by understanding the root cause of defects.

We have to work closely with industry in these areas to understand their business and concerns and we have developed and transferred technologies and applications to SMEs and multinational companies that can be potentially scaled up to manufacturing companies in Ireland and internationally.

What are some of the biggest challenges you face as a data analytics researcher?

I would say the main challenge is getting industry to see the importance or effectiveness of our research. In general, research is only effective when it is applied to processes that are changed considerably enough to have an impact. When it comes to analytics, most of our projects are based on data that companies may neither have nor use. Thus, getting the right data and showing the value in it is the first step, and the most important one, to get companies involved.

It is like trying to convince someone to change their habits knowing that they have always worked. The only way to do so it is by showing them with numbers and facts that there is always room for improvement.

Are there any common misconceptions about data analytics research?

The most common misconception about data analytics is that it is all about building ‘magic’ models. In fact, building a model is a single layer within the multiple phases that comprise a data analytics project.

For instance, delivering an analytics project includes data collection, data cleansing, verification of the data, data visualisation, exploratory data analysis, etc. At the end, all these steps can take up to 90pc of the time invested in the project with the other 10pc in developing the model.

Another common misconception is that analytics only works in large organisations with large resources. Analytics projects do not require sophisticated infrastructure to process and get the most value out of the data. In contrast, what matters is how data is used and interpreted to extract valuable information that will benefit business practices. This and the belief that access to large volume of data produces better and more accurate results often become barriers when engaging with small and medium enterprises. The best way to address these myths is by showing companies what they are capable of with the resources they have.

What are some of the areas of research you’d like to see tackled in the years ahead?

I would like to see where the concept of smart manufacturing will head. Will industry 4.0 and the technologies developed in it be embraced by the manufacturing sector? I think that is a question that will take a long time to answer, especially if we consider the technological differences and acceptance rate between the countries. I believe the adoption process is something industry and academia have to closely work together on if we want to make the most of it.

And, personally, now that I have seen the impact it can have, I would love to see some of the technologies adopted by Irish industry – such as collaborative robots, augmented reality or artificial intelligence – being implemented in Colombia.

Reporting: Silicon Republic