SingPost CIO on what makes supply chains smarter

Logistics play a pivotal role in navigating the complex headwinds of global trade. Image created by DALL·E 3.
- Advertisement -

The boom of e-commerce has had an immediate impact on the logistics industry, necessitating not only timely deliveries but also speeds surpassing expectations.

This is where technologies such as generative AI, coupled with a well-designed cloud strategy, are pivotal in enabling organisations like SingPost to excel in their core function — overcoming geographical barriers.

Noel Singgih, Group Chief Information Officer of SingPost, sat down with Frontier Enterprise to talk about his logistics experience, the tech outlook for the industry, and how SingPost intends to maintain its competitive advantage amid a changing enterprise landscape.

You’ve been with DHL for about 10 years before you moved to SingPost in July last year. So how’s the transition been, from a global logistics giant to SingPost?

I think there are two parts to the story. The first is the distinct organisational cultures at SingPost compared to DHL. However, the commitment from the management committees – including the group CEO – in driving tech transformation, is key to sustaining positive, long-term change.  This culture makes it similar to a global organisation as well. This isn’t just for the sake of doing it, but to show that Singapore Post isn’t just a traditional postal company that people think of.

Noel Singgih, Group Chief Information Officer, Singapore Post. Image courtesy of Singapore Post.

The second part is about the technology transformation. As we harness technology for transformation, this is not about the complexity of figuring out what to do, because it’s clear what needs to happen. The complexity is more about how to manage the changes so we can deliver every aspect — from forming that foundation, introducing new capabilities, and aligning that back with the business strategy, to ultimately delivering value to stakeholders.

Eighty-five percent of our revenue comes from outside Singapore, indicating that our business is not just postal; it reflects that we are really a global logistics company. Nearly 50% of the revenue coming in is from Australia, where we are one of the top 4PL (fourth-party logistics) providers. Then, there’s 15% from Singapore, and the rest is from our international cross-border business, where we connect merchants to consumers worldwide. This is made possible because of our relationship with postal networks and our commercial segment, where we use air cargo and other commercial means.

So, it’s a big difference, kind of taking you back to past experiences, but also teaching you what you should and should not be doing.

In terms of technology, what transformations are you currently looking at?

E-commerce is currently our main growth driver and I believe it will remain key for the foreseeable future. But to capitalise on this effectively, a few technological aspects need to be addressed. One is the ability to collect and manage the data. It’s important to appreciate the fact that the supply chain industry is quite fragmented, involving various layers like the first mile, last mile, and so on. Managing all the data exchange and handling the seasonal volume spikes in e-commerce, which can increase tenfold during certain weeks – then drop again – requires flexible technology capacity.

When I joined SingPost, we were operating on a hybrid system with a bit of on-premises technology, although most of our system was already cloud-based. By now, we have everything in the cloud. We’ve decided to go fully public cloud, securing it and making sure that cybersecurity and data protection are central in our control centre. We are also engaging partners who align with our sustainability agenda.

Is that multiple public cloud services?

Yes, we work with multiple cloud providers, but we have a preference for one that has already achieved carbon neutrality. We’re partnering closely with Google Cloud mainly because of their lead in carbon neutrality, which aligns with our sustainability goals. This partnership is part of our initiatives to reduce carbon emissions. We’ve also found that Google Cloud’s expertise and technologies are particularly helpful in enhancing our technology and business capabilities.

But it’s important to note that we’re still working with other providers like AWS, Azure, and AliCloud. We’re currently in the process of harmonising our operations to determine where the bulk of our cloud activities will be focused on.

How do you deal with disparities in the supply chain, for example, when some of your partners are not digitalised yet?

I think managing the complexity in supply chains is key to profitability. We still need to handle a lot of unstructured data, in written or printed formats, that needs to be integrated into our system for effective global supply chain management. Currently, we have a team manually entering this data, and this is part of the Google AI Trailblazer program we are doing. We’re developing AI prototypes to convert and make sense of the data in scanned documents, despite coming in different formats.

We are trying to identify crucial details like shipment values, shipment origins, and the presence of dangerous goods in the shipments. This allows us to pre-empt or pre-warn as the data is being consumed into the system.

Currently, our process involves people manually reading through these documents. You can imagine the difficulty, as some of these documents, even though printed, are not in the best conditions. Words are printed on top of the lines, some are fading, and sometimes even the human eyes have difficulty reading them. However, from the prototypes we’re developing, we see a massive opportunity. These could enable the system to alert our operations team about specific concerns, such as identifying dangerous goods in the air waybills we process. This is a significant advancement, helping us to focus on more value-added activities.

What are some other AI use cases SingPost is looking at?

We are looking at how we can use generative AI for our internal business processes. One area that we’re looking at is customer services. Initially, we’re looking at how generative AI can help train our agents. Given the high turnover typically seen in customer service roles and the need to keep up with frequent internal updates, we believe generative AI could act as a support tool for our agents.

The second area we’re looking at is responding to requests for proposals (RFPs) or RFx. We’ve noticed that many RFx queries are asking the same things, just phrased differently. We’re exploring how generative AI can assist in identifying effective responses, ensuring compliance, and spotting any inaccuracies in how we respond to these RFx queries.

The third area we’re looking at is how generative AI can help our lawyers in contract reviews. Specifically, we want to see how it can help our lawyers quickly identify clauses or statements that might violate our internal ‘golden principles.’ If such issues are flagged early, we can focus on addressing them quickly.

Which technologies on the horizon do you see impacting logistics and supply chains the most?

I think the focus will still be around data, and how data will be used. I will link that back to generative AI. For generative AI to work effectively, you need good quality data. It’s also about collecting more and more data. I believe this is where IoT and computer vision come into play. We are piloting computer vision to identify unsafe behaviours in our operations. For instance, if someone is walking in our warehouse or sorting centres without safety vests, or outside the pedestrian zones, computer vision will spot and flag it. Safety is paramount for us, hence the pilot. But imagine computer vision being used to identify damaged packages. Integrating this data into our overall orchestrating engine will obviously be helpful.

I think digital twins are going to be significant as well. The ability to bring all this data together and visualise it in a way that you can almost see the operation in real time will be instrumental in how we simulate and plan for various scenarios.