- Home
- Covid-19
- Fighting the COVID-19 pandemic
- Chapter 5: Staying ahead
Chapter 5: Staying ahead
Discover how the SAF's Medical Operations Task Force used data analytics and technology to coordinate COVID-19 patient care across facilities, track 180,000 migrant workers, and predict healthcare demands during the pandemic.
On this page

COA MG Goh Si Hou and CDF LG Melvyn Ong being briefed on how Medical Operations Task Force improved situation awareness by fusing multiple sources of epidemiology data and healthcare capacity data.
Source: MINDEF
Introduction
Towards the end of April 2020, things seemed to be looking up. The daily number of new cases among the general public seemed to have peaked with the introduction of the “Circuit Breaker” in early April and progressively tighter control measures in the following weeks.
By that time, EHTF and JTF(A) had swung into action. Four SAF camps had begun operating as temporary housing facilities for migrant workers and a few more camps were in the process of being converted for the same purpose. Every day, about 400 migrant workers from at-risk dormitories were being processed to move into these clean sites and other isolation facilities. But the problem was far from being over. COVID-19 cases continued to soar in the dormitories and demand for care facilities and resources was surging. More agencies had to be roped in to provide the care. JTF(A) was responsible for recovery-in-place in the dormitories; MOH managed the hospitals and CCFs; many others ran the Community Recovery Facilities (CRFs) and worker decant sites. Everyone understood that the mission was to isolate and care for the patients. But a key element was missing. There was no coordination to piece together information on the COVID-19 patients and project requirements for the beds in hospitals, CCFs or CRFs. The centralised entity in MOH assigning hospital bed spaces nationwide was a small outfit with limited capacity. Moreover, each facility had its own bed management system. With no overview of the flow of patients, many care facilities came close to being overwhelmed.
Management of Covid-19 patients based on medical condition

In the early days when COVID-19 struck Singapore, confirmed, suspected and “clean” cases were managed via targeted approaches. Many different facilities were set up for different purposes such as testing, isolation and treatment. The large number of cases required painstaking tracking and coordination, as well as resource-intensive transportation.
There was also the challenge of how to efficiently transport patients from one facility to another. Without centralised control, the emergency transportation (or conveyance) system was on the verge of being overloaded. There was an urgent need to develop comprehensive situation awareness of the stock and flow of COVID-19 patients across the care facilities and to set up a centralised emergency transportation system. The SAF task forces also had to look beyond these immediate needs and anticipate operational demands so that we could gain the initiative in the fight against the virus.

SAF personnel supporting the conveyance of migrant workers across recovery and isolation facilities.
Source: MINDEF
Setting up the Medical Operations Task Force (MOTF)
The need for an integrated operations hub was clear. So MOTF was set up to support MOH in carrying out three key tasks. The first was to develop a “single source of truth” on the numbers and movements of COVID-19 patients, by aggregating the data from multiple sources to a single, reliable source. The second was to coordinate the allocation of medical resources and transport operations across all entities. The third was to anticipate operational demands to minimise the risk of being overwhelmed by an unpredictable, deadly enemy. In order to carry out these tasks, MOTF would have to develop technology solutions, as part of a larger ops-tech initiative, to track and monitor COVID-19 cases among migrant workers as well as resource utilisation rates.
The Assistant Chief of the General Staff (Intelligence) [ACGS(Int)] COL Tan Cheng Kwee was given command of MOTF. Personnel from HQ Army Intelligence (Army Int) and G2-Army staffed this task force because of the analytical nature of its mission. With the need to operationalise the task force as quickly as possible and to work closely with MOH, MOTF quickly set up shop in the Health Promotion Board (HPB) building near MOH. Associate Professor Raymond Chua, Deputy Director of Medical Services (Health Regulation Group) of MOH, joined MOTF as Deputy Commander.

A team comprising officials from the SAF and MOH briefing Senior Minister and Coordinating Minister for National Security Mr Teo Chee Hean on the set-up of the MOTF.
Source: MINDEF
A team from 6 Div who had served in HSTF was brought to the MOTF Secretariat to quickly establish MOTF's battle cycles, operating processes, and links with adjacent agencies to open the information channels. HQ Army Int and G2-Army personnel made full use of the information flows to immediately start sense-making. Personnel from JTF(A) joined the MOTF's current operations cell to plan daily transport operations.
Within a week of starting work, MOTF had ironed out the ownership of resources and every component of the entire transportation process chain with JTF(A). MOTF was now giving the SAF better situation awareness and enabling tighter coordination of effort and sharper concentration of force in critical areas.
Taking stock and regaining control
COL Tan co-chaired daily MOTF meetings with MOH's Director of Medical Services, Associate Professor Kenneth Mak, to direct the operations. At the beginning, the top priority was to address the lack of a comprehensive overview of the flow of COVID-19 patients. MOH needed an accurate and timely situation picture to make policy decisions. MOTF knew that it had to present the information in a way that was easy to understand and helpful to the decision makers.
The team worked hard to build the Operations Dashboard. This was something military personnel were familiar with and they quickly established a frame, pulled data from the hospitals, dormitories, and recovery and isolation facilities, and put the dashboard to work. The task force knew they were not going to get it right the first time. “Honestly, at first the margin of error was quite high. But we nailed it down before long,” COL Tan recalled. A mere week and a half after it was operationalised, the very first iteration of the dashboard received positive feedback from the Cabinet and Multi-Ministry Task Force. This dashboard would be an important step to regaining control of the COVID-19 situation in the migrant worker dormitories.

From left to right: CPT Sabian Lim from HQ Army Int, Mr Chung Wai Kong from DSTA, and MAJ Henry Lee from Systems Integration Office. They developed a command and control system which helped agencies to plan and coordinate the medical care for migrant worker COVID-19 patients.
Source: MINDEF
MOTF case flow operation dashboard

The MOTF case flow operational dashboard provided a comprehensive overview of the daily caseload and work flow from multiple sources of information.
Source: MINDEF

SAF and DSTA personnel who developed technology solutions in the MOTF's Ops-Tech Centre.
Source: MINDEF
Data collection - "Rubbish In, Rubbish Out"
MOTF understood only too well the basic principle of data management - rubbish in, rubbish out. It had to make sure that the inputs were as accurate as possible so that the dashboard would be functional and useful. This was a mammoth task with many operational challenges. Initially, the task force did not even know where to pick up migrant worker patients to transport to CCFs and CRFs. It was difficult to locate them because many had been moved out of their registered dormitories. Although their employers had to update the registered residence within 14 days, this was far too slow for MOTF to convey them to the right care facility in a timely manner. Clearly, there had to be a system to track COVID-19 patients, and clean up the data so that MOTF could act on it.
Taking a leaf from the Army's experience when it moved its tracking of NSmen through numerous nominal rolls to more digital means, MOTF quickly shifted to a similar digitalisation strategy with the Migrant Worker COVID tracker. This was designed so that every individual's data and movement could be easily updated, even if he could not do it himself. MOTF explored various ideas, from wristbands to QR codes, and eventually settled on the Migrant Worker Conveyance Monitoring (MWCM) mobile app. This tracked the movement of a COVID-19 infected migrant worker by scanning the barcode of his Foreign Identification Number and updating the system on his location and status of conveyance.

The Migrant Worker Conveyance Monitoring mobile app.
Source: MINDEF
The MWCM app was deployed to all Swab and Isolation Facilities, Government Quarantine Facilities, CCFs and decant sites. The system was also integrated with apps such as MOM's Tenant Management Module, which captured information on a migrant worker's move into and out of a dormitory.
Unfortunately, the MWCM app was not widely used at first by managing agents at the facilities so there were gaps in information on the workers' location. To encourage managing agents to use the app, the team added onboarding and monitoring features that managing agents would find useful. They followed up closely to get regular feedback from the managing agents and provide assistance. Eventually, the MWCM app became the one-stop data collection method that simplified and organised COVID-19 data inputs for the migrant worker population.

Migrant worker health status and movement details are entered into a web-based app and the information is consolidated at the Operations Dashboard automatically for planning and tracking purposes.
Source: MINDEF
The agile development and scaling of digital solutions such as the MWCM app was possible because of the strong ops-tech partnership between the SAF and DTC. Knowing that technology would be a key enabler to carry out its mission, MOTF had brought in DSTA engineers from the start to form the nucleus of its ops-tech team. DSTA spearheaded the digital system architecture and design with operational input from the SAF, and reached out to industry partners such as ST Engineering and NCS Pte Ltd to develop, integrate, deploy and maintain the systems. This strong ops-tech partnership generated tailwind for MOTF in its race against the clock to safeguard the national healthcare system from being overwhelmed.
Data fusion and Sense-making - Identifying the "Single source of truth"
With the data collection processes in place, MOTF had to figure out a way to incorporate the data in the digital platforms that the task force was using. This took dedication and hard work by the personnel in the newly formed Data Fusion Centre (DFC). They started work at 0800 hrs to prepare for the first huddle of the day at 1000 hrs - which gave them only two hours to “close all the accounts” from the previous night and provide an updated situation picture.
The operational challenges of this task were daunting. The data came from multiple sources, with different time-stamps and sometimes in conflict with other data. Some were raw data while others had gone through some form of processing. As the centres had different operating times, there were time-lags in data processing and always people in transit. The various databases had to be synchronised, so the team had the arduous task of assessing all the data for reliability and accuracy. Lags and discrepancies had to be accounted for and loose ends reconciled in order to establish a “single source of truth” that would be useful for making decisions and formulating policy.
MOTF's data cycle

To meet the vital need to synchronise the databases, DFC had to have a good grasp of the various data cycles, a data cycle being the sequence that each unit of data goes through from the time it is generated to the time it is applied. As a central database, DFC had to fuse data from a large number of stakeholders and integrate more data cycles in order to provide an accurate and coherent situation picture. The different data cycles of DFC's partners had to be aligned with a single common reference point, and timestamped to indicate when the data was collated and represented.

The MOTF Data Fusion Centre daily operations.
Source: MINDEF
It was gruelling work. The data had to be processed manually, and the team often worked into the wee hours of the morning, before starting a new day of work just a few hours later. But within two months, it had managed to develop an automated solution and dashboard which presented an overview of the entire COVID-19 medical operations. With this “single source of truth”, the various task groups in MOTF could now operate together seamlessly. This dashboard was also helpful to other task forces and agencies for their planning and operations. Whenever there were data discrepancies, DFC would seek clarification from the relevant agencies. This helped to uncover causal factors and derive corrective actions.

A Data Fusion Team having a quick pow-wow at the start of the day.
Source: MINDEF
Data-based projections - Moving ahead of the curve
The success of the dashboard gave MOTF confidence to go even further and look beyond managing the situation day to day. It quickly developed projections to predict chokepoints and excess capacities so that it could forecast supply and demand for task force operations. For instance, to optimise the movement of COVID-19 patients, it developed a matrix to prioritise admission to medical and care facilities, taking into consideration factors such as the patient's medical risk profile. Using the projections, MOTF also worked with JTF(A) to adjust the testing schedule so that potential spikes in the load on medical facilities would be better managed.
The data projections also helped in strategic resource planning. At one point, when MOH was considering adding more CCFs and CRFs for migrant workers, MOTF's projections convinced MOH to increase the number of recovery sites instead. This helped avert a chokepoint at CCFs. Allowing the migrant workers who were medically well but still tested positive to move to facilities with leaner medical staff meant that precious medical resources could be diverted to care for more seriously ill patients.
180,000 migrant workers were conveyed across 153 COVID-19 facilities nationwide.— -


Data scientists in MOTF analysing data to aid timely decision-making.
Source: MINDEF
Indeed, data proved to be the new gold. Data fusion and data-driven decision making were critical for the success of MOTF's operations. The experience showed that with data fusion and sense-making elements, especially in operations involving multiple nodes, decision makers can have a clearer situation picture and tighter operational control. This enables them to make better decisions. For protracted operations, data can be trended over time to make projections. This is helpful for strategic planning and enables some level of prediction of how courses of action will pan out.
The experience in the COVID-19 fight showed that there must be clearly defined roles for the different groups of people who deal with data so they know what is expected of them. There should also be a clear structure for meaningful data-driven decision making. It comes down to competencies in managing data, and data managers are just as important as data scientists. The MOTF team realised that it was not prudent to only have data scientists cleaning data. After the data is fused, data managers are required to manipulate the data and extract insights. Proper roles, responsibilities, and ops-tech structures must be established upfront to effectively harness the benefits of data.
The SAF must recognise the power of using data and harness it to solve problems as operations and missions become increasingly more complex. We have to improve our data literacy, get comfortable with making use of data, understand the challenges around the use of data and develop the structures to make effective use of data. Data analytics is one key tool that SAF leaders must be able to exploit.