Data drives systems by providing the foundation for decision-making, optimising processes, personalising experiences, enabling predictive analytics and facilitating continuous improvement. It plays a critical role in today’s interconnected and data-rich world, allowing organisations to operate more efficiently, make better-informed decisions, and deliver enhanced experiences to their users or customers.
Colonel Shubhodoy Mukherjee, CEO of Mumbai Metro One, spoke with ET Edge Insights on a variety of topics, including how data drives Metro systems in Mumbai, data security, and data mining.
Edited excerpts
What does data mean for an organisation such as Metro One, and how does data drive a system like Metro?
First, let me explain how data benefits Mumbai’s Metro riders. With data, we improve our efficiency and, consequently, the train’s performance. The second most essential factor is ensuring the safety of all our passengers. Thirdly, we make sure that the comfort of our customers is at the forefront of all we do.
The data collected from the systems, which include the train system, the signalling system, and the automatic fare collection system, contributes to efficiency. This vast amount of dynamic data gives us a complete picture of the system’s health, as well as its loading, journey details, the number of trips that should be made in the event of a sudden influx of passengers, and whether we should increase the number of trains for safety reasons. The trains operating on the track are controlled by an automated system known as the automated train protection system.
To elaborate on the safety factor, even though trains are three minutes behind each other, this is the entire layer of protection that is ensured within the system, and it is accomplished by a constant exchange of data that is picked up from the track to see the distance between one train and the next and to see the correct voltage coming into the train to power it.
Commuters are the third important factor. By accumulating data, we can determine which fare products best suit them, as well as their travel patterns and so on. Thus, we construct our entire data-driven system there.
These three elements constitute the essential backbone of our Metro system.
How does your company encourage a data-driven culture among its employees? How do you, as a leader, fit into this plan?
Before we can begin mining data for insights that will improve our system and the commuting experience for everyone involved, we need to determine what information is most valuable to you and how best to make use of it. Therefore, from the beginning, we instructed individuals to examine all data from the perspective of the system.
For instance, on the structural safety issue, we need to monitor the Cable Stay Bridge across Western Express Highway, which is the first of its kind in India. This bridge is monitored through a very advanced array of sensors that provides us vital data on the vibrations and stresses, which helps us carrying out Predictive Maintenance as against Preventative or Corrective Maintenance, which in this particular instance could be a very costly proposition.
On a similar but different data set requirement, we often examine which stations are the most frequented in terms of commuter behaviour? How do we incentivise this? Which of their products are being used more frequently, and which of their products are less utilisied? Which items should be removed from the system? Which product should we develop and promote and so on.
To effectively work with data, it is crucial to have a clear understanding of which data is relevant, how to present it, and most importantly, how to extract valuable insights from it. Thankfully, our technical workforce is quite youthful with an average age of 24. For this generation, consuming data is second nature, and as a result, the entire team adopted it with relative ease. So we are lucky in that way.
Where my leadership comes into play is that I have brought that feature back to the forefront of the leadership team’s minds.
How do you extract useful information from vast datasets? Do you have a strategy for data mining?
Data overload is the greatest trap, and if those dealing with data are not sensitised to what data they need to extract, it will result in meaningless data that you cannot use for insights or as decision enablers. We determine, on a scale of one to five, which decisions have an effect on our systems and commuters. We draw these conclusions based on the observed impacts. We look at the detailed data we have collected and see what kind of data really works. Then we make the datasets, and these datasets help us make these important decisions.
For example, in the past, Monday and Friday were the busiest days for commuting. Now, Wednesday and Thursday are the busiest days due to the rise of hybrid work, post Covid 19. On the basis of this information, we are developing a dynamic timetable that accommodates the demand on Wednesdays and Thursdays.
So it has been a continuous learning process because nothing remains the same.
What steps have you taken to safeguard your data from unauthorised access or breaches?
The nerve centre for all our data resources is our Operation Control Centre which is completely secure and the data room has a huge sanctity. There are very few authorised people who can access it. The access to the data is usually controlled. Then we have a built-in redundancy. We have an alternative control centre.
Then there are firewalls in place, the very severe serious laterals in place. But beyond that, the important thing is the question of privacy, of data, of individual data, because we’re dealing with public data. Therefore, we attempt to anonymize public data to the greatest extent possible so that we can analyse commuter behaviour. And no data, by any chance, is linked with the personal inputs that we wash out of the system before we bring them to the forefront.