Community projects
Building the city together, solving the problems we all share
Gridlock every evening rush hour, meals you cannot be sure are safe: these worries are shared by millions of people, and no single organization can solve them alone. New Era is working on AI solutions to these problems and invites experts, businesses, public authorities, investors and residents to build them, test them and bring them into everyday life together.
The projects
01 · Urban traffic · Ho Chi Minh City
Smart traffic coordination
The city does not lack roads. It lacks a brain to coordinate them.
-
13.3 million
vehicles, including more than 1.5 million cars
Ho Chi Minh City Traffic Police, 10/2026
-
12.8%
of urban built-up land is used for transport; the 2026–2030 target is 16–26%
Department of Construction · Resolution 489/NQ-HDND
-
+24%
rise in congestion in a single year
Ho Chi Minh City Department of Construction, 12/2025
-
~USD 6 billion
lost to traffic jams every year
Department of Transport and Ho Chi Minh City Institute for Development Studies, 2022
Vehicles are multiplying faster than roads can be built, and that gap cannot be closed with concrete.
Breaking a habit of thought
“Build wider roads and the jams will go away.”
Experience around the world shows the opposite. Research by Duranton and Turner in the American Economic Review (2011) found that traffic grows in proportion to the roads added. Texas spent more than USD 2.8 billion widening the Katy Freeway in Houston to 23 lanes; three years later, the morning commute took 30% longer and the evening commute 55% longer.
Traffic works like a computer. The road surface is the hardware; the way traffic is coordinated is the software.
Powerful hardware with buggy software still freezes.
1,070 signalized intersections in Ho Chi Minh City, early 2025
In Hanoi, one month after AI cameras were connected to the traffic lights, travel times on the pilot routes fell by 20–36.8%, without adding a single metre of road.
Five lines of research
Rewriting the software of the streets
-
Multi-phase signals that change with the time of day
Morning peak, midday, evening rush and night each get their own signal plan. Neighbouring intersections are linked into green waves so traffic flows through in one stretch instead of stopping at every junction.
-
Connecting signals to the control centre over satellite internet
Each intersection controller gets a satellite link to the traffic control centre: it can be installed right away where no cable has been laid, and it keeps the connection alive when a cable fails. Low-orbit satellite internet has been licensed for a pilot in Vietnam.
-
Smart traffic diversion with matching signs
When a corridor starts to clog, the system works out a detour in advance and shows it on electronic guidance signs a few junctions earlier, so drivers turn off before they reach the jam.
-
Sensors that measure density at every hour
AI cameras and sensors count the flow, measure queue lengths and tell motorbikes, cars and buses apart. The data is updated continuously for every intersection and every hour of the day.
-
Electronic road signs and an AI operating system
No-entry, lane, one-way and two-way signs switch electronically by the hour, like the reversible lane on Cong Hoa Street in operation since 5/2026. The AI operating system gathers every signal onto a live map of the city and proposes how to coordinate traffic; human operators review and decide.
Roadmap
-
Data research
Collect traffic flow data and pick one typical congested corridor.
-
Simulation
Build a digital twin of the corridor and test each coordination plan on the computer before touching the real road.
-
Pilot
Run it on a real corridor and measure travel time, number of stops and queue length before and after.
-
Scale up
Extend it across the whole city and to other large cities.
Who we need on board
- Traffic engineers
- Data and AI specialists
- Traffic signal and sensor manufacturers
- Telecom and satellite companies
- Urban traffic authorities
- Residents sharing data on jam spots
Data sources
VOV 29/10/2026 (Ho Chi Minh City Traffic Police) · Dai Bieu Nhan Dan (Ho Chi Minh City Department of Construction) · Resolution 489/NQ-HDND of 10/12/2025 · Znews 10/12/2025 (Ho Chi Minh City Department of Construction) · Dai Doan Ket 17/07/2022 · Duranton & Turner, American Economic Review 101(6), 2011 · City Observatory 16/12/2015 (Houston TranStar data) · Phap Luat TP.HCM 10/01/2025 · Bao Dau Tu 13/01/2026 (Hanoi Police) · Decision 659/QD-TTg (VnEconomy 26/03/2025) · Tuoi Tre 15/05/2026
02 · Food safety
Handheld food scanner
Spot suspicious food right at the market, within minutes.
-
~134
food poisoning outbreaks a year, nearly 3,800 people affected and 24 deaths
Ministry of Health, average 2011–2025
-
~32,000
hospital admissions a year from foodborne disease, far more than the cases reported
World Bank, estimate
-
573
fake milk powder brands in one ring, sold even to patients and premature babies
Ministry of Public Security, 4/2025
-
7–10 days
to wait for laboratory results on one suspicious sample
Ministry of Health, 7/2026
The one-week gap
The results arrive a week later. The fresh food was sold long ago.
The rapid test kits in use today detect only a few substances such as borax, formalin and dyes. Suspicious samples must go to a laboratory and wait 7–10 days; by the time the results come back, the vegetables, meat and fish are already on the dinner table. Meanwhile fraud keeps getting more sophisticated: more than 4,100 pigs were seized for being raised with banned substances in Dong Nai (2026), and about 2,900 tonnes of bean sprouts soaked in a growth stimulant were sold in a single year (Dak Lak, 2024).
The idea
Every substance has its own light fingerprint. A palm-sized spectral scanner shines light on a piece of meat, a fish or a pack of milk and records the reflected spectrum. AI compares that spectrum with a spectral library of Vietnamese food and sends the result to a phone within minutes, with no cutting of samples and no chemicals.
-
Pass
-
Suspicious
needs further testing
-
Fail
Already tested: handheld spectrometers have measured the freshness of salmon, tuna and beef with 85–92% accuracy (Sensors, 2020) and recognized fish treated with chemicals to look fresh (Foods, 2024). The US Food and Drug Administration (FDA) uses handheld spectrometers to screen imports arriving by international mail; the European Union funded the PhasmaFOOD project to develop a food scanner connected to a phone.
Three groups of users
-
Residents
Quickly check meat, fish, vegetables and milk at the market or supermarket before buying.
-
Businesses
Canteens, schools, restaurants and supermarkets check every incoming batch and keep a history for each supplier.
-
Public authorities
Screen widely at wholesale markets, slaughterhouses and border gates, sending only suspicious samples to the laboratory. Results flow into one shared database for early warning, in line with Politburo Conclusion 114-KL/TW (11/2026).
Clear about the limits
The scanner is a fast, wide screening layer; findings of violations and penalties still belong to the laboratory. It works well for freshness, counterfeits, adulteration and chemical treatments at high concentrations. Pesticide residues at very low levels and disease-causing bacteria still need a laboratory. The project’s job is to establish exactly what the scanner can do for each Vietnamese food, and to say exactly that.
Roadmap
-
Spectral library
Collect samples of Vietnamese food with matching laboratory results.
-
AI models by food group
Meat, fish, vegetables, milk, spices: one model for each group.
-
Pilot
Canteens, supermarkets, wholesale markets; compare scanner results with the laboratory.
-
Scale up
Bring the scanner and the warning data to residents, businesses and public authorities across Vietnam.
Who we need on board
- Analytical chemists and food scientists
- Reference laboratories
- Spectrometer manufacturers
- Canteens, supermarkets and markets for pilots
- Food safety authorities
- Investors
Data sources
Ministry of Health, Report 1244/BC-BYT of 06/7/2026 · World Bank, Food safety risk management in Vietnam, 2017 · Tuoi Tre 12/4/2025 (Ministry of Public Security) · Tuoi Tre 05/8/2026 · VTC News 31/12/2024 · Sensors 2020, DOI 10.3390/s20154299 · Foods 2024, DOI 10.3390/foods13213443 · Spectroscopy Online 12/10/2023 · CORDIS, project 732541 · VietnamPlus 09/11/2026
Join
Everyone has a part to play
A better city does not come from one organization. It comes from the knowledge of experts, the resources of businesses, the permission of public authorities and data from residents’ own daily lives.
-
Experts and scientists
Contribute specialist knowledge, challenge the plans and help design the trials.
-
Businesses
Contribute equipment, technology and trial sites, and help bring the products to market.
-
Public authorities
Set the problem, share data and open the way for pilots.
-
Investors and sponsors
Provide financial support from research through to scale-up.
-
Residents
Share data from daily life, try each version and give feedback.