Page 35 - AI for Good-Innovate for Impact

AI for Good-Innovate for Impact

Use Case – 4: NET-PEAT-ZERO: Networked ASEAN Peatland

Forest for Net-Zero 4 - UPM

Country: Malaysia

Organization: Universiti Putra Malaysia (UPM)

Contact person: Aduwati Sali aduwati@ upm .edu .my

4�1� Use case summary table

Domain Forestry

Problem to be Decision-making by forest rangers, firefighters and the local community
addressed on potential peatland forest fire could trigger transboundary haze in the
ASEAN region. The early warning system is influenced by varied factors
such as climate conditions, soil temperature, soil humidity and ground-
water level. The problems include water level in the peatland soil, water
management, machinery, environmental sustainability, forest rangers, fire-
fighters and local community’s training and education, market access, and
government policies.
Key aspects of Unifying the data of various agencies and using machine learning models
the solution to predict the best plans, policies, and strategies to increase crop produc-
tivity and economic growth.

Technology IoT, Open data, mobile apps, AI/ML and geospatial analysis.
keywords

Data availability Private

Metadata (type Text data
of data)

Model Training kNN
and fine-tuning

Testbeds or pilot https:// doi .org/ 10 .3390/ fire6070272
deployments https:// doi .org/ 10 .3390/ f14071472

https:// hal -emse .ccsd .cnrs .fr/ emse -03346490

4�2� Use case description

4�2�1 Description

The tropical area has a large area of peatland, which is an important ecosystem that is regarded
as home by millions of people, plants and animals. However, the dried-up and degraded

19