The InnoLeap programme seeks to connect public sector
agencies with challenging problem statements and Research Institutes (RIs), Institutes
of High Learning (IHLs), and commercial entities with innovative solutions. It
offers participants an opportunity to ideate and co-create innovative digital
solutions to address the needs of agencies.
The submission of solutions closes on 15 March. More details
are available at: https://ideas.ecitizen.gov.sg/egp/process/EGOV/EideasChallenge/?challengeId=20069.
The problem statements are as below.
of compound to detect safety incidents and compound defects
The team receives feedbacks of faulty facilities or illegal
activities taking place in the user’s outdoor compoung outside working hours.
This results in extra hours being put in to carry out inspections at night or
after working hours. The user is seeking potential Artificial Intelligence (AI)
and robotics solution to patrol the outdoor compound during non-office hours
based on various pre-programmed routes.
The solution should incorporate video/image/sensor analytics
to detect safety incidents and compound defects, such as smoking, speeding,
unleashed dog, pot holes and faulty lights. The detected incidents have to be
alerted to Ops Centre and alerts should be accompanied by relevant original image/video/data
and information on location of incident.
AI solution to
provide concierge and visitor assistance
To enhance productivity of existing manpower resource and
visitor experience, user is seeking an AI solution to provide concierge and
visitor assistance in an outdoor open compound. The solution is needed at designated
fixed location within the compound.
The solution should be able to comprehend and answer general
enquiries posed by visitors, in natural language (preferably multi-lingual). It
should be able to print map and routes to help visitors get around the compound.
Upon request from a visitor, it should be able to call for taxi service and
contact the Operations Command Centre directly via intercom.
Take photographs of
tall objects like buildings and statues easily
It is physically challenging to photograph tall objects like
statues and buildings (height from 2 to 12 metres) for photogrammetry . It is desired to leverage on
digital technologies to take photographs of tall objects to obtain information
of the subject (i.e. dimensions – basic length, width and height).
information on historic building
It has been challenging to crowdsource information
especially on historical or heritage content and attempts have not usually garnered
much public interest.
The user is seeking ideas and implementation (e.g. social
experiments) to create engaging and interesting ways to collect information on
historic buildings that will help to preserve and celebrate the shared heritage
of Singapore’s diverse communities for the purpose of education,
nation-building and cultural understanding.
The collected information on historic buildings would enrich
existing database and potentially be used for analysis and assessment of
heritage sites affected by urban redevelopment. It could also be curated into
cultural heritage stories for the general public
The user is looking for a safe and interactive Mixed Reality
solution, preferably to incorporate AI/ machine learning for educating and
training users, where their unique nature of operation works (e.g. interviewing
skills) and situations are simulated in the virtual environment.
Current training methods are conducted in a conventional
manner – eLearning videos and role play with props setups supplemented with 2D
photos. This lacks an immersive real-world experience and different scenarios
can only be executed in role plays to display various forms of dynamism.
Risk Profiling of
An effective framework for licensing healthcare institutions
requires a good assessment of the risk profile which can then be used to guide
inspection approach before issuance of license.
The user is interested in exploring digital and AI solutions
to support a risk-based approach to regulate the healthcare industry.
The idea is to build an assessment system based on a dynamic
and robust analytics model to aid in risk profiling of licensees and recommend
inspection and enforcement actions relevant to the needs of licensees.
The assessment system
should support: 1) generation of decision trees, checklists and reports to
guide decision making (e.g. inspection actions) prior issuance of license; and 2)
full data query, data mining and intel reporting tool to meet the reporting
needs of user.
The assessment system will eventually be integrated into the
user’s enterprise licensing system. In view of an evolving healthcare landscape,
the system should continue to grow smarter and relevant with more data to
support and refine its analytics model.
Auscultation refers to listening to the internal sounds of
the body, usually using a stethoscope. It is an important clinical tool for
clinicians to support diagnosis of lung and heart conditions and it plays a key
role in the management of patients in nursing homes, or who are bed bound at
When a patient with cardio vascular disease (heart problem),
chronic obstructive pulmonary disease (lung problem) or respiratory infection
is discharged from the hospital, there is no solution available to monitor
their heart and lung sounds at home. So critical events may be missed or there
might be unnecessary hospital or emergency department visits due to false
Therefore, the user is seeking a Smart Auscultation system
or solution that will greatly benefit telehealth initiatives for the elderly
patients where constant access to medical staff is limited. Such a Smart Auscultation system or solution
could be powered by AI that listens and classifies sounds to aid clinicians in
making a decision.
 The American
Society for Photogrammetry and Remote Sensing defines photogrammetry as
the art, science, and technology of obtaining reliable information about
physical objects and the environment, through processes of recording,
measuring, and interpreting images and patterns of electromagnetic radiant
energy and other phenomena. The input to photogrammetry is photographs, and the
output is typically a map, a drawing, a measurement, or a 3D model of some
real-world object or scene