Southeast Asia (SEA) is a well-connected, mobile-first region with a future. Smartphone penetration is expected to double in the next five years in Indonesia, Myanmar and Philippines, and already exceeds 100 per cent in Malaysia, Thailand and Singapore.Aware of the potential of huge troves of data now being collected by people leaving a digital footprint, governments are working together with the region’s O2O (online-to-offline) companies to harness the data and apply deep-learning techniques. Their stated aim is to provide income opportunities along with services with added safety and convenience that the new middle class has not previously found with the region’s more traditional institutions.

Ditesh Gathani, head of engineering and leader of Grab’s efforts to employ machine-learning, data science and artificial intelligence to provide reliable transport and mobile payment solutions for its customers, was recently interviewed by The Nation. We discussed the current situation of on-demand transportation and the region’s digital economy. 

Why is Southeast Asia the fastest growing region in the world for on-demand transportation? 

According to the “E-conomy SEA” study by Temasek & Google (May 2016), the online rides industry is expected to grow by five times by 2025, reaching US$13 billion (Bt419 billion) and making up 15 per cent of the total $90 billion travel market in this region.

In Thailand, online ride-hailing is expected to become an around $1.9-billion market, growing at 20 per cent over the next 10 years, making Thailand the second-highest ride-hailing market after Indonesia. This enormous potential means that Grab can contribute to society through our leading transport and mobile payment solutions.

We have one of the region’s largest data sets and combine it with top-tier machine-learning, data science and artificial intelligent (AI) capabilities. In 2017, we have hit 1 billion rides and we are now doing 3.5 million rides daily in 155 cities in SEA. More than 2.1 million driver-partners earn income through our platform. More data means better service for drivers, passengers and cities, 10 terabyte of data generated each day, making up multi-petabytes of insights over the years.

One example of how this data helped is our estimating time of arrival calculations. Most map providers aggregate data across all vehicle types as they do not differentiate vehicle types.

Every ride on our platform is logged to improve the algorithm according to the needs of each consumer in each city. [For example], because of congestion a Bangkok GrabShare rider is used to a bit more waiting time than a Singapore GrabShare rider.

We believe we can play a role in shaping what SEA’s cities will look like 10-20 years down the road and have heavily invested in our data science, machine learning and AI capabilities.

Today, we have around 60 top-tier people on our data-science and data engineering teams. In the short term, we optimise our supply and demand today, offering drivers a way to earn more, faster; offering passengers the safest, most affordable and most convenient mode of transportation.

We invest in our platform and its data-capabilities, not just for profit. This is the fastest and most effective way to improve the way our drivers, passengers and cities live and work – moving us closer to fulfilling a vision of a ‘smart city’ under Thailand 4.0 policy.

How can Southeast Asia accelerate the online-to-offline (O2O) transformation?

More than 75 per cent of people in Southeast Asia have no bank accounts. Moreover, the emerging middle class in SEA does not trust banks and the classic financial infrastructure. Almost nine in 10 of all transactions in SEA happen in cash and only one in 10 people has a credit card.

At the same time, Southeast Asia has the ‘digital opportunity gap’. Every month, in SEA, 4 million people come online for the first time, and the number of online consumers will jump from 260 million today to 480 million in 2020.

This new generation of consumers does not use laptops and does not trust old financial institutions, but they use their mobile for all kinds of services.

The digital economy offers more opportunities today to live a better life than ever before, but there are still significant barriers and divides and that means not everyone can access them. Just like in transport, it’s the lack of safety, convenience and simplicity that prevents people from benefiting from the digital economy. Grab as a platform solved for these challenges in transportation and is now tackling these same issues for payments.

So O2O players will drive change in this space. Entrepreneurs and consumers can leverage on mobile technology to plug into the digital economy. For example, it is already empowering members of our GrabPay agent network to be entrepreneurs and boost their incomes even in the more far-flung areas of Indonesia – no need for payments terminals and formal bank accounts, all you need is a smartphone.

Mobile-first adoption will continue to drive a truly on-demand business model [and] businesses and merchants who understand this and mobile consumer behaviour can tailor services and products to capitalise on this trend.

How does Grab integrate O2O transformation strategies into the business? 

We are uniquely positioned for success in O2O because we are already one of SEA’s largest marketplaces with 1 billion transactions annually. Transport is the most important use-case for [going] cashless in Southeast Asia today and Grab is SEA’s leading ride-hailing platform.

Our investments in the last 12 months have started a virtuous cycle for our O2O platform: our large user base of 72 million mobile phones, in 155 cities and eight countries already find GrabPay easy to use.

In the coming months, our strategy is to make GrabPay available everywhere to everyone.

This means using Grab without cash needs to be as easy as it is with cash. It also means we are working with partners to make it as easy to pay for their services as it is to pay for a GrabCar or GrabTaxi.

How does Grab utilise machine-learning, data science and artificial intelligence to provide reliable transport and mobile payment solutions via the Grab platform?

Our data science team analyses real-time and historical demand data to alert drivers to anticipated demand hotspots. Our internal study conducted in June 2017 shows Grab driver-partners in Thailand on average earn 19 per cent more than the national average wage per hour after joining Grab.

For our passengers, our technology continues to make the service safer and more convenient.

Grab’s telematics initiative tracked whether drivers are driving safely: speeding, hard braking, sudden acceleration and dangerous swerving.

After sending reminders to drivers who were speeding, there was a 35 per cent reduction in speeding incidences across Southeast Asia. Traffic accident rates with Grab are on average five times lower than the country average.

We also anticipate the needs of our passengers on the platform by learning our passenger’s favourite location and pre-filling their destination. For example, if you go to Silom Road for work every Tuesday, but not most other days, the system can predict your drop-off point and pre-fill the destination in your app on Tuesday as soon as you open it.