Challenge Statements
The Horizon
Arena for bold ideas, blue-sky thinking and future-forward concepts
000
Current State:
Currently, NPPs/NPCs support transactions that do not require direct involvement of officers, such as depositing found items and reports lodging. There is a need to review self-help kiosk's technologies, accommodate bulky items in drop boxes, and address the misconception that they are manned 24/7 as some turn to them during emergencies.
Aim:
Reimagine how our NPPs/NPCs can be transformed into friendly, intelligent, autonomous hubs that can interact with and efficiently serve the public 24/7, even when no human officer is on site.
Currently, NPPs/NPCs support transactions that do not require direct involvement of officers, such as depositing found items and reports lodging. There is a need to review self-help kiosk's technologies, accommodate bulky items in drop boxes, and address the misconception that they are manned 24/7 as some turn to them during emergencies.
Aim:
Reimagine how our NPPs/NPCs can be transformed into friendly, intelligent, autonomous hubs that can interact with and efficiently serve the public 24/7, even when no human officer is on site.
Current State:
Firefighters conduct fire suppression and search-and-rescue operations in extreme conditions involving intense heat, smoke-filled environments, and structurally compromised buildings. Their operational effectiveness depends heavily on their gear and equipment, where enhancements in technologies for thermal regulation, mobility, and situational awareness will significantly boost their operational capabilities.
Aim:
Design a next-generation firefighter gear and equipment, to enhance the safety of our firefighters while they battle the harsh elements and navigate within Singapore’s complex urban terrain to save lives.
Firefighters conduct fire suppression and search-and-rescue operations in extreme conditions involving intense heat, smoke-filled environments, and structurally compromised buildings. Their operational effectiveness depends heavily on their gear and equipment, where enhancements in technologies for thermal regulation, mobility, and situational awareness will significantly boost their operational capabilities.
Aim:
Design a next-generation firefighter gear and equipment, to enhance the safety of our firefighters while they battle the harsh elements and navigate within Singapore’s complex urban terrain to save lives.
Current State:
Prison transporters are custom-built vehicles designed to transport inmates within and outside of prisons. Protected by metal cabins, reinforced windows, secured compartments, with remote tracking system, prison transporters ensure safe and secured custody of our inmates on the road. While effective, these transporters need a refresh to meet SPS' evolving operational needs.
Aim:
Revolutionise our prison vehicles for the digital age through smart technology to boost transport efficiency and to ensure inmates' wellbeing while upholding the highest levels of security.
Prison transporters are custom-built vehicles designed to transport inmates within and outside of prisons. Protected by metal cabins, reinforced windows, secured compartments, with remote tracking system, prison transporters ensure safe and secured custody of our inmates on the road. While effective, these transporters need a refresh to meet SPS' evolving operational needs.
Aim:
Revolutionise our prison vehicles for the digital age through smart technology to boost transport efficiency and to ensure inmates' wellbeing while upholding the highest levels of security.
Current State:
Police officers in Ground Response Force and Traffic Police undergo advanced driving and riding training beyond basic courses, focusing on emergency manoeuvres. There is a need to review training methods to meet modern operational needs, enhance specialised skills, implement objective competency assessments, and maintain safety standards despite existing challenges.
Aim:
Transform driving and riding training for our Police into an immersive, technology-powered experience that enables officers to confidently navigate high-stakes, real-world scenarios on the road.
Police officers in Ground Response Force and Traffic Police undergo advanced driving and riding training beyond basic courses, focusing on emergency manoeuvres. There is a need to review training methods to meet modern operational needs, enhance specialised skills, implement objective competency assessments, and maintain safety standards despite existing challenges.
Aim:
Transform driving and riding training for our Police into an immersive, technology-powered experience that enables officers to confidently navigate high-stakes, real-world scenarios on the road.
Makers' Zone
Hardware prototyping where innovative products are shaped by hand
000
Current State:
Conventional robotic systems with pre-programmed motions are inadequate for chemical spill clean-up tasks. This limitation stems from the complexity and unpredictability of operational environments, coupled with the multi-stage nature of clean-up procedures to ensure completeness and safe disposal. To this end, imitation learning has some advantages over classical techniques.
Aim:
Train a robot to clean chemical spills expertly using vision-based sensors through imitation techniques, where a robot learns by observing an expert's actions or behaviour.
Conventional robotic systems with pre-programmed motions are inadequate for chemical spill clean-up tasks. This limitation stems from the complexity and unpredictability of operational environments, coupled with the multi-stage nature of clean-up procedures to ensure completeness and safe disposal. To this end, imitation learning has some advantages over classical techniques.
Aim:
Train a robot to clean chemical spills expertly using vision-based sensors through imitation techniques, where a robot learns by observing an expert's actions or behaviour.
Current State:
Most iris biometric authentication solutions today rely on bulky hardware sensors integrated into electronic devices. This significantly limits the portability of the solution, making them less practical for mobile deployment in dynamic operational environments. However, reducing sensor size typically results in decreased image quality and lower authentication reliability.
Aim:
Design a pocket-sized iris scanner to be used with a mobile device for self-authentication. The scanner may fit over the device’s existing camera or be connected via cables.
Most iris biometric authentication solutions today rely on bulky hardware sensors integrated into electronic devices. This significantly limits the portability of the solution, making them less practical for mobile deployment in dynamic operational environments. However, reducing sensor size typically results in decreased image quality and lower authentication reliability.
Aim:
Design a pocket-sized iris scanner to be used with a mobile device for self-authentication. The scanner may fit over the device’s existing camera or be connected via cables.
Current State:
Hostile Autonomous Vehicles (AV) can pose threats to public safety and critical infrastructure. These vehicles typically employ multiple redundant sensing systems, such as LiDAR, cameras and radar. An effective counter measure could be to target their suite of sensors to safely halt or redirect rogue AV without causing collateral damage.
Aim:
Design a smart system with the ability to disrupt the navigation system of dangerous autonomous vehicles without affecting other vehicles or the surrounding smart infrastructure.
Hostile Autonomous Vehicles (AV) can pose threats to public safety and critical infrastructure. These vehicles typically employ multiple redundant sensing systems, such as LiDAR, cameras and radar. An effective counter measure could be to target their suite of sensors to safely halt or redirect rogue AV without causing collateral damage.
Aim:
Design a smart system with the ability to disrupt the navigation system of dangerous autonomous vehicles without affecting other vehicles or the surrounding smart infrastructure.
Current State:
Most methods for sealing indoor spaces during airborne chemical threats are slow, unreliable, and potentially destructive on surfaces, like using duct tapes and plastic sheeting for windows and vents. A simple, reusable system is needed to quickly create airtight seals on openings, without causing damage or requiring extensive restoration.
Aim:
Design an innovative “snap-and-seal” emergency kit that transforms any room into a safe space within minutes – like creating an instant “clean room” in the house.
Most methods for sealing indoor spaces during airborne chemical threats are slow, unreliable, and potentially destructive on surfaces, like using duct tapes and plastic sheeting for windows and vents. A simple, reusable system is needed to quickly create airtight seals on openings, without causing damage or requiring extensive restoration.
Aim:
Design an innovative “snap-and-seal” emergency kit that transforms any room into a safe space within minutes – like creating an instant “clean room” in the house.
Code Core
Fast, fluid and scalable, where ideas go digital and software-driven solutions are born
000
Current State:
The key to protecting financial systems and national infrastructure lies in the timely identification of cybersecurity threats showing up as anomalous activities in high-volume datasets. AI offers potential for efficient anomaly detection, but the challenge lies in the vast amount of normal data versus that of sparse anomalous data.
Aim:
Build an AI guardian for cybersecurity, trained on normal data and non-malicious activity records, to spot and explain anomalies in a complex dataset – akin to an investigator in the digital space.
The key to protecting financial systems and national infrastructure lies in the timely identification of cybersecurity threats showing up as anomalous activities in high-volume datasets. AI offers potential for efficient anomaly detection, but the challenge lies in the vast amount of normal data versus that of sparse anomalous data.
Aim:
Build an AI guardian for cybersecurity, trained on normal data and non-malicious activity records, to spot and explain anomalies in a complex dataset – akin to an investigator in the digital space.
Current State:
Haze in Singapore is primarily caused by transboundary air pollution from forest fires and exacerbated by seasonal weather pattern. Early satellite-based detection of such environmental threats plays a vital role in monitoring, forecasting, and mitigating its effects. However, different weather conditions may be difficult to distinguish from the top-down view.
Aim:
Develop a classification model that leverages computer vision to accurately differentiate haze and smoke threats from normal clouds and other environmental elements in satellite images.
Haze in Singapore is primarily caused by transboundary air pollution from forest fires and exacerbated by seasonal weather pattern. Early satellite-based detection of such environmental threats plays a vital role in monitoring, forecasting, and mitigating its effects. However, different weather conditions may be difficult to distinguish from the top-down view.
Aim:
Develop a classification model that leverages computer vision to accurately differentiate haze and smoke threats from normal clouds and other environmental elements in satellite images.
Current State:
Hazardous chemical plume remains an ever-present threat due to industrial developments. To accurately assess these threats, urban population data is used. However, footfall traffic of nature trails and parks are not widely studied in the context of public safety, which renders the harm assessment for such locations challenging.
Aim:
Build a tool that helps predict when and where people gather in parks, so help can reach them quicker, in the event of an emergency.
Hazardous chemical plume remains an ever-present threat due to industrial developments. To accurately assess these threats, urban population data is used. However, footfall traffic of nature trails and parks are not widely studied in the context of public safety, which renders the harm assessment for such locations challenging.
Aim:
Build a tool that helps predict when and where people gather in parks, so help can reach them quicker, in the event of an emergency.
Current State:
Documenting evidence at expansive crime scenes is a time-consuming task. Crime scene specialists currently rely on manual methods to mark and record the position of placards, often transferring these locations by hand onto sketch plans. When the crime scene is expansive, so does the required complexity and effort.
Aim:
Create a smart tool to help officers record forensic evidence and their locations faster and more accurately, enabling them to conduct investigations while on-the-go.
Documenting evidence at expansive crime scenes is a time-consuming task. Crime scene specialists currently rely on manual methods to mark and record the position of placards, often transferring these locations by hand onto sketch plans. When the crime scene is expansive, so does the required complexity and effort.
Aim:
Create a smart tool to help officers record forensic evidence and their locations faster and more accurately, enabling them to conduct investigations while on-the-go.