‘Disruptive’ Artificial Intelligence to Make Jakarta More Livable

2 May 2019

Jakarta. Indonesia’s often chaotic capital Jakarta has been implementing artificial technology in the past few years to help it cope with flood, traffic, security issues and waste management problems with few successes.

Artificial intelligence (AI) technology has helped officials anticipate floods earlier than before, the head of Jakarta Smart City’s planning, research and development unit Alex Siahaan said on Tuesday.

“We’ve experienced fewer floods in the past few years. This is partly because officials have been able to respond to flood warnings faster. Using early information we get from AI technology, officials can act before the floods hit. Floods will still happen, but technology makes us more efficient [in anticipating them],” Alex said.

The Jakarta administration has collaborated with Nodeflux, a Jakarta-based AI startup, using its people-counting technology to analyze people’s movement in often crowded areas, such as a bus stop.

“The governor [Anies Baswedan] believes the key to [driving] industry 4.0 in Jakarta is collaboration. He wants collaborations not only inside the administration, but also with startups and residents. He also wants to improve security in Jakarta, making it not only a ‘smart city’ but also a ‘safe city,’” Alex said.

Nodeflux is also trialing its license plate recognition technology to help increase the city’s tax revenue in collaboration with the Jakarta administration through its Jakarta Smart City program.

“We want to create, explore and introduce disruptive technology to Indonesians. It’s a blessing in disguise that we live in Indonesia, a country with a lot of problems to solve such as traffic, waste and floods. It’s a playground for us to create solutions using technology. We want to implement technology that can solve problems around us,” said Faris Rahman, Nodeflux’s chief technical officer and co-founder.

Nodeflux’s most popular product at the moment is its facial recognition technology.

“Telkom’s [Indonesia’s largest telecommunications provider] subsidiary Infomedia uses our facial recognition technology in their office building,” Faris said.

The head of NVIDIA at Binus University’s AI research and development center Bens Pardamean also believes technology should be used to solve social problems

“One of my missions, and also Faris’ at Nodeflux, is to start building Artificial Intelligence communities in Indonesia, to create a government-supported program called AI Indonesia. Our model is AI Singapore, which is also supported by the Singaporean government,” Bens said.

Nodeflux has also worked with the National Police to provide surveillance at high-profile international events such as the 2018 Asian Games and the 2018 International Monetary Fund and World Bank annual meeting in Bali.

Nodeflux is backed by Indonesia’s largest telecommunications company Telekomunikasi Indonesia and venture capital East Ventures.

Nodeflux’s product line includes:

  • Face Recognition: automatically detects facial features and classifies them into “known person” and “unknown person.” It can be used to identify customers and people who have been put on a blacklist.
  • License Plate Recognition: reads numbers and characters on license plates of any vehicle, even in bad weather, poor lighting and even if the license plate has been modified.
  • People Counting: detects human traffic in a certain area. It can be used to count the number of people visiting the area and identify its peak hours.
  • Vehicle Counting: counts every passing vehicle and classifies them into motorcycles or small, medium and large vehicles. It can classify up to 23 types of vehicle.
  • Vehicle Dwelling: identifies and calculates how long a vehicle stops at a certain spot. It can be used to monitor illegal parking and manage traffic.
  • Crowd Behavior: detects behavior anomalies like running and fighting. It can be used to safeguard an area by spotting sudden movements in a crowd.
  • Vehicle Intrusion: detects vehicles entering restricted areas. It can be used to spot violations and send alerts.
  • Water Level Detection: measures water level in dams, sewers and waterways. When water volume exceeds normal level, the system will immediately alert on-duty officials.