Startup ideas: wildfire forecast, better customer communications

November 16, 2020 09:10
FlameMapper uses artificial intelligence to construct models for preventing and managing wildfires. Photo: FlameMapper

It is estimated that the global population was less than 1 billion at the beginning of the 19th century. Today, 200 years later, the world’s population has soared to 7.2 billion, and is expected to reach 10 billion by 2050. As a result, humans have to compete with all other creatures for natural resources. Human activities have caused global warming which caused the world many problems. One of them is the drastic decrease in worker productivity due to hot weather, the other is serious wildfires. In 2019, Australia’s bushfires caused 17.1 million hectares of scorched earth (about 155 Hong Kong), and 1 billion animals suffered from the disaster. In 2020, the wildfires on the west coast of the United States killed dozens of people, the smoke even spread to Europe, which is 8,000 km away.

As long as the issue of global warming has not been adequately addressed, the frequency and scale of wildfires will only increase in the coming days. The troubling issue has driven startups to find solutions, such as FlameMapper, a member of the Esri Startup Program (Esri Program).

The two founders, one old man and one youngster, have both suffered from wildfires in the past. The 10-day fire in Malibu, Los Angeles, California destroyed more than 300 houses, including the 8-year-old Shea Broussard’s home. Although the elder Anthony Shafer's home remained intact fortunately, the stigma is in his mind.

The startup was founded in 2015, using artificial intelligence to prevent and manage wildfires. It helps planning how to protect infrastructure, provide evacuation routes, and pinpoint high-risk locations in case of wildfires. When the wind direction changes on the spot, the prediction model can be adjusted in near real time to help firefighters, government, and public organisations, including power companies and schools to determine the priority of assigning resources. According to the startup company, these forecasts can help to protect important public facilities and reduce the extent of damage by 70%.

In the event of a major incident like wildfire, apart from disaster relief, it is also important to ensure close communication between relevant units and the public, in order to guide all parties to cooperate and avoid panic. DataCapable, founded in 2013, also a member of the Esri Program, is enabling these rescue efforts.

It collects and analyses data extensively, and uses machine learning to detect special incidents, such as gas leaks, wildfires, earthquakes, power outages, school shooters and more. It can often send out alarm to customers and emergency centre operators 40 minutes earlier than traditional event awareness workflows and systems.

For example, Seattle City Light, one of the top 10 public power companies in the U.S., serves over 400,000 residential and commercial customers. Once a power failure occurs, it will certainly cause much customer dissatisfaction. DataCapable sorts out the comments on social media, and immediately reports the dissatisfied messages to the power company for handling as quickly as possible. At the same time, when and where these messages were initiated, as well as relevant information such as demographic characteristics of the district, customer behaviour data, and natural environment are all displayed on a map dashboard, so as to assist the company in making decisions.

In addition, DataCapable provides utility companies with real-time power outage reports, customer sentiment feedback (how many people are expected to be dissatisfied), weather data (indicating power outage possibility) to help improve service satisfaction. In fact, FlameMapper and DataCapable are business partners, they complement each other to solve problems for customers.

-- Contact us at [email protected]

Adjunct Professor, Department of Computer Science, Faculty of Engineering; Department of Geography, Faculty of Social Sciences; and Faculty of Architecture, The University of Hong Kong