We are creating some awesome events for you. Kindly bear with us.

New Solutions from HK Lab

Recently, two tech firms operating under the Hong Kong Smart Government Innovation Lab announced that they have launched new solutions which are now ready to be acquired by companies and institutions.


Solution description

The firm’s solution is an AI-Cloud wildfire detection SaaS that detects forest/outdoor fire that uses standard video surveillance cameras to capture images, then, Machine Learning is applied for objection detection on the images to identify fire and smoke.

The solution is automatic and extremely effective at detecting forest and/or outdoor fires. With the current climate change crisis and the ongoing pandemic, there is a growing number of wildfires as more people are visiting country parks and hiking trails.

The AI-Cloud wildfire detection SaaS helps:

  • Protect valuable natural resources like forests, country parks as well as rural areas from wildfires;
  • Reduce the carbon emissions caused by unwanted forest fires; and,
  • Mitigate wildfire risks and damages to nature, communities, and citizens.

Application Areas

The solution was developed to be applied in the areas of Environment and Infrastructure.

Technologies Used

The solution employs the latest in Artificial Intelligence (AI) as well as Cloud Computing.

Use case

The solution is able to:

  • Detect wildfires and mitigate wildfire risks and damages
  • Detect wildfires and reduce the carbon emissions from unwanted forest fires
  • Detect fires that may occur on construction sites, thus improving construction work safety
  • Detect fires near infrastructure (such as in an electricity distribution network) and minimize fire risks and damages and services provided to the public
  • Detect fires near rural residential areas to protect lives and assets


Solution description

The firm has developed intelligent energy monitoring systems that are simple and easy to deploy. Data is collected and transmitted to a cloud platform and analysed with the firm’s AI model, which is trained and has been proven to have accurate predictions.

It dynamically accounts for the influencing factors of energy usage (e.g., temperature, humidity and seasonality) to predict energy usage and determine a budget.

Real-time energy data with granular zoning, i.e., down to the AC usage in a meeting room, is visualized in the software with easy-to-understand and quantifiable metrics so users know if they have overused or conserved energy.

This is also facilitated by a series of customized notifications, interpersonal energy training workshops, and through the use of automation, the firm is able to raise energy conservation awareness and knowledge, hence driving behavioural change intelligently in communities.

Application Areas

The solution was developed to be deployed in the areas of City Management, Environment, Finance as well as Housing.

Technologies Used

The solution employs the latest in Artificial Intelligence (AI), Cloud Computing, Data Analytics, Deep Learning, Internet of Things (IoT) and Mobile Technologies.

Use Case

  • User-friendly software platform derived by 8000 users to enhance the awareness and understanding of energy and encourage behavioural change
  • Real-time and granular energy usage monitoring and notification to spot abnormality and take prompt actions
  • Various data analysis metrics by day, by week, by month or by AC, light and socket to deduce energy pattern and saving opportunities
  • AI modelling for energy budget with consideration of actual usage and environmental factor provide realistic target and quantifiable tracking of energy performance for users
  • Energy comparison of different buildings/zones at customized period provide easy benchmarking in a single glass of plain
  • Establish energy usage intensity (EUI) metric in kWh by meter sq. to compare energy usage of own location vs location in the other similar building types
  • Scalable and agile infrastructure which can expand with further integration with other smart sensors or controls
Send this to a friend