When we moved into the second chapter of our ongoing Issue 01: Smart Cities, we titled it “Intelligent Island”. The phrase’s lofty assonance points to an aspirational vision of the total computerisation of Singapore, and the myriad Information Technology (IT) initiatives, masterplans and policies that were implemented since the 1980s until today. In broad strokes, here is a brief outline of our recent technological history, as the city was shaped into smartness by the Singapore Government:
A high-level ministerial Committee on National Computerisation (CNC) was set up to make recommendations to the Government how Singapore could benefit from IT. This was called the National Computerisation Plan, the first of several masterplans that have provided five and ten-year roadmaps for the nation.
One of the CNC's recommendations was to establish a National Computer Board (NCB), and so it came to fruition to coordinate computer education and training in the civil service with the Civil Service Computerisation Program (CSCP), and in the computer services industry.
These preliminary efforts were increased through the launch of the National IT Plan (NITP). Its Working Committee comprised of the NCB, the Economic Development Board (EDB), Singapore Telecoms, and National University of Singapore (NUS). The Plan’s objectives were to enhance Singapore’s competitiveness in boosting IT manpower, infrastructure, application, culture, etc. and to create a climate for creativity and entrepreneurship.
The TradeNet and PortNet websites were set up to facilitate documentation for the trading, logistics and shipping industries.
A high-level ministerial Committee on National Computerisation (CNC) was set up to make recommendations to the Government how Singapore could benefit from IT. This was called the National Computerisation Plan, the The NCB kickstarted the IT2000 study of more than 200 executives in the public and private sectors to examine how IT could be leveraged to grow eleven different industries from manufacturing to healthcare and media.of several masterplans that have provided five and ten-year roadmaps for the nation.
The NCB was reorganised and merged with the Telecommunications Authority of Singapore (TAS) to become part of the Infocomm Development Authority of Singapore (IDA), and given a new mandate to make the IT2000 findings a reality and to position Singapore as a global IT hub. Singapore ONE (One Network for Everyone) was launched to build the first nation-wide broadband network, involving the NCB, the TAS, the National Science and Technology Board (NSTB), the Singapore Broadcasting Authority (SBA), and the EDB.
Part of the NCB responsible for implementing the CSCP was corporatised to make it more efficient. It became the National Computer Systems(NCS). Today the NCS is a member of the Singtel Group.
A new millennial blueprint called Infocomm 21 was announced with the aim to develop Singapore into a thriving infocomm capital. The e-Government Action Plan I (eGAP I) spearheaded by the IDA rolled out infocomm technologies and electronic services throughout the public sector, so that a total of 1600 public services were made available online.
eGAP II followed up on augmenting service excellence with the SingPass authentication system, Wireless@SG for free broadband access in public spaces, and the roll-out of 3G mobile technology. This was all steered by a revised masterplan called Connected Singapore.
A ten-year plan called Intelligent Nation 2015 (iN2015) leading up to Singapore’s 50th year of independence envisioned “An Intelligent Nation, a Global City, powered by Infocomm”. IDA launched iGov2010, which took the prior eGAP plans further to increase synergy and whole-of-government integration between the public sector and citizens.
The Smart Nation initiative was launched to develop Singapore into a smart city, and was set up in the Prime Minister’s Office.
The IDA and the Media Development Authority were restructured into the Infocommunications Media Development Authority (IMDA) and the Government Technology Organisation (GTO).
The IMDA announced the Digital Government Blueprint (DGB) to better leverage data and harness new technologies in support of Smart Nation. Its slogan was “A Government that is Digital to the Core, and Serves with Heart”.
Around the very beginning of this timeline , in the late 70s or early 80s, a tiny island-within-an-island called Pulau Saigon went missing. The so-far team sat down with artist-technologist Debbie Ding and the former Director of the Data Science and Artificial Intelligence Division of GovTech, Liu Feng-Yuan, to discover their mutual interests in the margins and blips of data collection. Through Ding’s own accidental, but nonetheless obsessive excavation, she recounts what was oddly forgotten in this race to computerise the city.
Adeline Setiawan: Let’s begin with a bit of memory work. Feng-Yuan, can you bring us up to speed with a short history of GovTech and what your role was in the agency?
Liu Feng-Yuan: Yeah! Back in the old days, when people were first setting up computers, it used to be called the National Computer Board (NCB). Then it got split up and some part of it got privatised to become National Computer Systems (NCS), the IT vendor. The other part became the Infocomm Development Agency (IDA), kind of an old-fashioned name. They used to do all these things with cabling, the promotion of technology, tech support for the Government, regulation, and just everything to do with IT.
Over time, these agencies evolved and shifted as the landscape changed. The media regulator, the Media Development Authority (MDA), and the technology regulator IDA realised that they were regulating the same people, like Singtel. That’s why they joined MDA and IDA together to become IMDA — great branding! [Laughs] And the bit about Smart Nation and technology became GovTech.
At the beginning of GovTech, we used to outsource a lot of work. Then we hired our own programmers to build apps and technology.
Debbie Ding: Why did you outsource at the start?
Liu: It was felt that civil servants wouldn’t naturally have the expertise, and that the private sector would be better at the technology. And then we reached a point where technology was so core to what we did, that if we didn’t have our own guys, then we couldn’t be a smart buyer of technology or do things in an iterative and agile way.
It started with the software team that built the MyResponder app for the Singapore Civil Defence Force (SCDF). If I had a heart attack and you guys had the app, then you could locate me and give me CPR.
Ding: And it had all the CPR and aid devices located as well.
Liu: Yes... Before, we couldn’t code or “clean” the data — couldn’t explore it or excavate it. That was my motivation in setting up the data team, and hiring people who could code and programme. I started running it to help our politicians make decisions with data. I’m not very ideological, but I’m obsessed that when we make decisions, there must be a process — it must be data driven.
It was also an experiment in doing things a slightly different way. Designers and engineers are very independent-minded, so it was a challenge to get them to work within this larger, more slow-moving bureaucracy. Although, don’t get me wrong, bureaucracy has its functions! If you want something to work for a large number of people, 95% of the time you need a system, which is how you have bureaucracy. But I worked with this team that provoked things differently.
And with Smart Nation, my own motto was “data-driven policy-making”. Smart Nation is really an exhortation to use technology in all parts of delivering public services, and for Singaporeans to embrace technology. But that’s a big challenge to define what stuff you should do in support of Smart Nation.
One thing my team built was this parking app  where you pay for public parking coupons on an app. Is it a really big Smart Nation initiative? Probably not — it’s a really simple thing. But it makes your life more convenient so I don’t have to run off in the rain to my car to pay for my parking. I think it was something my colleagues and I were kind of struggling with — how do you determine the functions and needs of the smart city. If I were to be really practical about it, it’s just better convenience. And maybe that's good enough.
Christina J. Chua: It’s also a question of what you do with, or how do you apply the data. We wanted to go back even further, maybe around the time when the NCB was just formed, to hone in on a marginal area of Singapore’s geography where data was especially scant. Debbie, can you share more about how your research on Pulau Saigon began?
Ding: The project began because I had been very interested in mapping the Singapore River. I had made an interactive work Here the River Lies that was a big projection of the river, and if you swiped it, you could warp the shape of the river over time. It was extrapolated based on all the maps of the river, and imagining it through the course of urbanisation becoming straighter and straighter.
Initially, the river was not changed by public policy — there was no control. It was as if everybody owned it, and it was very random they didn’t dredge it. During the colonial time, all these British engineers were being sent over to help with the dredging but they all had malaria and died. It was a big problem! So it was all up to whoever lived there to alter the shape of the river.
I showed this work at a few places, then at the ArtScience Museum. There was a big Titanic show at the time and some curator decided it would be good to have a work about Singapore at the time of the Titanic. So they included the piece, and next to it, they asked someone from NUS Museum, to loan a cabinet of items that archeologists had dug up from the Singapore River. They were from Pulau Saigon. The cabinet was small and the items were so tiny — almost insignificant. I was really underwhelmed by the items.
Liu: What kind of items were there?
Ding: They had pottery sherds, but they also had things that were much more quotidian like a spoon, a buckle... Because of this cabinet, I became interested in researching Pulau Saigon and contacted the archeologist, Jennifer Barry. I wrote her a snail mail letter, and she replied sending me a copy of the catalogue  they had written back then, which had all the items catalogued and detailed. I also contacted another archeologist, John Miksic, who invited me to come and see the objects. When I went there, a lot of them were tiny or broken. They also had this whole bag of random stuff that were not so identifiable.
Liu: Sometimes you go to these tourist sites, and you expect everything to be grand, but actually it’s all quite mundane!
Ding: In fact a lot of archeology is mundane. When I was asking them about the objects, they were like, “I’m an archeologist, I’m not going to make up a story about what I’ve found because the thing is just really this broken bit here.”
And when I tried to find out more about the site, it’s not really recorded. The only public municipal record I found is that it was formally an island where there was an abattoir — that’s all really! It’s quite boring and there’s not much to say. There’s not much of any oral history either. If you talk to people in my parents’ generation, not many knew there was an island called Pulau Saigon. You can only see through the maps that there were two big metal bridges, some warehouses, factories... And you see it in the street directory up until 1970. From 1975 onwards, you can’t find it. And now, basically the CTE (Central Expressway) cuts through this area.
I wanted to write about these objects, but I found I couldn’t write because there was nothing to write about! Even in the research papers, there was a geography thesis about Pulau Saigon, but even then it was mainly a speculative thing that said something like, “I imagined the sound of pigs from the abattoirs...”
At the same time, I was very drawn to the NUS Museum. They have this room that is called the Archeology Library. It’s not a library of books. It’s just these sherds and broken pottery in the middle of the basement floor. And around it you have all these huge pottery wares from China, India — they’re full, whole items. And then you come to the Singapore section, where you have all the broken ones... So that was the beginning of The Library of Pulau Saigon.
Liu: I’m curious about this Un-forgetting Machine. What is it?
Ding: Because I was not really able to record or recapture the objects that were on Pulau Saigon, I thought that it could only be a kind of weird un-forgetting, where it’s not really remembering it, truly. All I had to be certain about was the names of the items they had found, but even then, they were names ascribed to the objects by the archeologists who had worked on the site.
I took the list of the most inconsequential things. Of course, the archeologists were most interested in all the pottery and all the clay objects. But I took the part at the end, when they were like, “Uhhh... What is this?” Furthermore, this is a very recent archeological site — post-18th Century, so is it even archeology? This is modern history!
So, I googled all the images for each object.
Liu: So you typed in “vase”, “spoon”...
Ding: And got all the spoon photos. Then I used all the outlines to make them into a 3D thing in AutoCAD. They were projected from 2D to 3D. Some things were not so recognisable, like the two pink fragments. This pink uranium graph kept coming up in my search. Also, the results were very time-based. What I searched for and what I printed in one round in March, was a bit different a few months later.
Liu: So there was some randomness in it.
Ding: Yes, I took some significant liberties in order to produce the things. The whole process was quite playful. It was obviously not a script. There was a lot of human intervention and tweaking. But I did want them to be all about the same scale — about 8 cm. Then I printed out two batches of objects using a 3D printer. But even by today’s standards, it’s not always the same. You’ve got to tune the machine and change the speed or change the temperature, and the result is slightly different.
Liu: I’m trying to understand it from your perspective — why was it important to have this programmatic approach? Although there was some random variation to it, but there was this underlying structure that you wanted.
Ding: Firstly, I wanted to make objects. And I thought to sculpt them myself by hand. But I just felt like I couldn’t — they wouldn’t be complete. I also wanted to leave it to the machine to be the craftsman, and almost co-make it with the machine. So it could be that this is the sum of all human understanding, if you think of all the things that could be tagged with a certain label.
Liu: And Google gave you access to this “sum of all human understanding”, to be able to synthesise it.
Ding: But even then, there’s a lot of distortion already. The idea is that when people look at the objects, they obviously don’t look like the thing that they imagined at all. It’s quite far from it. So there’s this gap. There’s a bit of tension between what I do as a designer, to say that we are closing the gap of an understanding from what you expect a thing is to the actual thing. But of course, as an artist, I want to show that the gap of knowing and naming a thing is very far apart.
Liu: This is how deep learning algorithms train and map a network. They take words and map the meaning of words into a two or three dimensional space. They plot the words, so that the words that are more similar in meaning come together. Husband and spouse would be close enough. And words with analogies lie on a horizontal line. So you’d have husband and male, and wife and female. But obviously it encompasses all the biases of this corpus that you would have, for example, certain female words are closely associated with nurse. These biases become encoded and embedded by using these algorithms for Google’s search. So it’s quite interesting that these algorithms help you encompass the essence of the world’s knowledge, but they also incorporate all kinds of biases.
Setiawan: In a way, the neural networks are trying to eradicate randomness, and yet from our perspective, we see this as a certain leaning and bias. In your work at GovTech, you must have come across a lot of small data that could be random and kind of noisy. How do you reconcile it?
Liu: Sometimes you can find patterns in small data. About two years ago, there was a breakdown in the Circle Line MRT and for 3 months, no one could figure out what was happening. Finally, I received a call and sent down two of my data scientists. In half a day, they were able to isolate the cause of this mysterious breakdown to a single rogue train that was going around the network and killing other trains as it crossed paths with it.
Ding: Yes, I read about the rogue train!
Liu: Exactly. And people assumed that we had used big data. But it was only 200 rows of data, which in this day and age is not a lot. It just took some really creative ways to visualise the data, look at the patterns, and we saw this zig-zag pattern in the faults that were happening. One of the guys had some insight that maybe it was a train of destruction on the other track that was causing this. But just looking at the data, you wouldn’t know that it was this hidden train that wasn’t breaking down, even as it was causing this path of dying trains. It was small data, but it wasn’t a linear conclusion. Perhaps it was just an “Aha!” moment in that data scientist’s brain where he connected the visual pattern and the possibility in the real world. Data are just neutral fragments of information, and getting to know where it’s useful or productive is something we are trying to do.
Ding: The data itself is actually really boring. It’s the data scientist that has to make that connection — to make sense of it.
Setiawan: That “Aha!” moment also has something to do with chance. Debbie, you have some of that in your research too.
Ding: Yes, in fact, when I came to the end of making The Library of Pulau Saigon, by sheer chance, I met the man who had found the site. He was an amateur archeologist who was in his 80s, and his name was Mr. Koh Lian Huat.
When I was installing the work at NUS Museum, there was a young intern who went to NTU (Nanyang Technological University) on her own to research more about the Singapore River. There she was, watching a video, when a professor walked past and asked her about it as that was his specialisation. She told him about my exhibition, and he said, “My father was the one who found the site!” He had been looking for Neolithic wares in 1988, but hadn’t found any on Pulau Saigon.
Finally, when I met Mr. Koh, he showed me how he had studied the objects. He had photographed them from different angles. And he had done all these weird sketches of the objects. They were literally little squiggles, as he tried to describe them in his own way. So coming to the end of my whole production process, it was quite wonderful to meet him.
But I also want to say that even though a lot of the things I make are about memory, I’m not nostalgic. These are not nostalgic projects. I’m more like a person who comes and sees how things have changed before I knew about their existence. A lot of it is recording to try and piece together these things, to formulate a system for understanding the fragments that were left behind. As an artist, that's what I do.
Liu Feng-Yuan is CEO and co-founder of BasisAI, an early-stage start-up enabling digital enterprises to build scalable and accountable AI. Prior to this he built a start-up within GovTech to harness data and AI for the public good, as part of Singapore’s Smart Nation initiative.
1 so-far assembled this timeline from a paper by Thompson S H Teo and Vivien K G Lim, "Singapore — an 'intelligent island' moving from vision to reality with information technology," Science and Public Policy, vol 26, 1 (February 1999): 27-36. We also referred to the various reports and masterplans prepared by the IDA and later the IMDA. Some of the more recent plans can be downloaded through the GovTech and IMDA websites: https://www.tech.gov.sg/media/corporate-publications/egov-masterplans and https://www.imda.gov.sg/sgdigital/digital-economy-framework-for-action
2 Parking.sg is a native app that allows you to pay for your public parking remotely. See more at www.parking.sg
3 Jennifer Barry, Pulau Saigon: A Post-Eighteenth Century Archaeological Assemblage Recovered From a Former Island in the Singapore River (Stamford: Rheidol Press, 2000). The catalogue details an archaeological excavation carried out between November 1998 and March 1989 where items made of ceramics, glass, bone, metal, wood, stone, plastic and rubber, as well as faunal and floral remains were discovered on Pulau Saigon.