SINGAPORE - Massive fires tearing through industrial warehouses often cause millions of dollars in damage and put lives at risk.

But forensic experts here think they may be able to predict such incidents and stop them from even happening.

Ms Lim Chin Chin, director of the Home Team Science and Technology Agency’s (HTX) Forensics Centre of Expertise, said HTX is looking at the predictive potential of forensics – especially when it is combined with data analysis and new technologies like artificial intelligence (AI) and machine learning.

She said: “By analysing data collected from past incidents and crime scenes, and finding the patterns and trends, we are looking at whether we are able to predict and prevent incidents from happening.

“Not just crime, but other incidents such as fires.”

Referring to chemical fires, she said past forensic data like burn patterns and types of chemicals involved can be analysed to identify trends, to prevent such fires from occurring.

She said: “By collating and analysing the data, we ask then if we are able to see certain trends, whether it usually involves certain chemicals or certain types of factories.

“We can then perhaps implement certain operating procedures, safety codes or guidelines to prevent such incidents from happening again.”

Forensics has traditionally been associated with crime scene investigations because of television dramas, which portray the field as something that revolves largely around fingerprinting and DNA collection.

But at HTX, the forensic experts led by Ms Lim are looking to develop Singapore’s capabilities in forensic science – not just to be reactive in solving crimes, but also to be proactive in preventing them.

HTX is a statutory board created in 2019 to develop science and technology capabilities for the Home Team.

In July 2022, the Washington Post (WP) reported that a group of University of Chicago scientists had unveiled an algorithm claiming to predict crime with 90 per cent accuracy.

The algorithm identifies locations in major cities that it calculates to have a high likelihood of crimes, like homicides and burglaries, occurring in the next week.

The software can also evaluate how policing varies across neighbourhoods in eight major cities in the United States. But WP reported that algorithms are only as good as the data fed to the system, and in the US, much of that data is biased against poor neighbourhoods and people of colour.

WP quoted criminal justice scholars, policing experts and technologists who said even if an algorithm was accurate, it could still be used by law enforcement to target people of colour and those living in poorer neighbourhoods for unjustified surveillance and monitoring.

In Singapore, HTX is using AI, modelling and simulation technologies to develop its forensic capabilities in more practical and reliable ways. For example, AI is used to predict and differentiate handwriting.

And computational science simulations are used to predict whether certain unknown substances have psychoactive effects and thus fall under the ambit of Singapore’s latest laws to tackle new psychoactive substances.

Ms Lim said while this is an area still in development, she hopes Singapore’s forensic technologies are able to advance further into the predictive space in the coming decade to solve crimes faster.

“We are not yet at the stage where we can know before someone commits a crime,” she said.

But if technology could be used to analyse, based on certain markers, like whether a weapon was brought to the scene, or where it originated from, she added, “we can solve crimes faster and increase the probability of catching the culprits, better protecting Singapore”.

Currently, crime scene specialists from the Forensics Division are deployed at the Singapore Police Force’s Criminal Investigation Department to support investigations by gathering evidence and reconstructing scenes.

They use tools such as blood screening kits that can test blood stains on site, an electrostatic lifting apparatus to collect footwear prints, and a portable fingerprint development chamber to develop fingerprints at the scene.

They also use special forensic lights to find hidden clues not visible to the naked eye, and 3D scanners to revisit the scene virtually.

As the Forensics Division is under HTX, its crime scene specialists can access resources across the agency to develop new solutions that make their work more efficient and can go beyond just solving crime.

One example is the use of mixed reality training which was developed by the Forensics Division and engineers from HTX.

The system focuses on analysing bloodstain patterns, allowing trainers to design virtual mock crime scenes and insert evidence, such as weapons, alongside different types of bloodstains.

Trainees then go through the scenario, marking out the relevant pieces of evidence virtually before their performance is evaluated by the trainer.

Drones are also being used for forensic crime scene documentation, with some forensic specialists trained to be unmanned aerial vehicle (UAV) pilots.

UAVs can be deployed at larger scenes to capture images from multiple aerial angles and create 3D reconstructions.

This not only helps cover larger areas, but also minimises any contamination of evidence.

But even with all of the technologies being developed and used, Ms Lim said the most crucial element in advancing the field is still people.

She said: “The field of forensics is very wide, and the development of technologies in the area is moving so fast.

“What we want now is to employ people with experience. We need people with industrial experience to join us to work on bigger projects that can reshape entire fields in science and technology.”