Students make devices to warn of damaged or flooded road surfaces

Research by Pham Van Phuc, Phan Dinh Nam, Pham Quang Trinh, Nguyen Van Thang (University of Technical Education) and Pham Phuong Nhi (University of Economics) at the University of Danang.

The group’s idea for implementing the project came from many tragic accidents that occurred due to damaged and flooded road surfaces. After spending more than a year of research, the team came up with a plan to use accelerometers and water level sensors with a compact design, low cost, easy to deploy and not affected by the weather. Two acceleration sensors are located along the rear wheel shock absorber. A water sensor is installed under the vehicle and a water level sensor is installed on the side of the vehicle to retrieve data.

When the vehicle is running on the road, the data collector will read the acceleration sensor, water level sensor, and water detection sensor. These data will be sent to the central processor via the 4G network.

The team uses an algorithm to determine road surface roughness through data from two acceleration sensors. Water and water level detection sensors help determine the level of flooding on the road. Data is sent and displayed on the phone screen in real time. To evaluate the system, the group experimented on urban roads in Da Nang such as Nguyen Chi Thanh, Le Loi, Tran Phu… for two months. Because the weather did not have heavy rain, the team simulated flooding to make the system work. Data on road surface condition and flooding is 95% accurate. The screen displays road surface condition in color, blue represents good road surface and turns orange and red for potholes from mild to severe. The flood level is also measured and warned by the system in centimeter units.

According to Pham Van Phuc, there are currently a number of technological devices in the country such as unmanned aerial vehicles (UAVs), integrated with cameras to process images, helping to determine the condition of the road surface. However, UAVs work well on terrain such as national highways, but are not suitable for smaller inner-city roads and hidden sections such as road tunnels. In addition, this technology depends on weather, lighting conditions, and rain flooding the road surface due to the use of cameras, leading to inaccurate results.

According to the evaluation team, the system still has many disadvantages such as in cases where the vehicle runs through the deceleration line, the system mistakenly reports that the road is damaged. The team’s solution is to apply artificial intelligence to train a model to distinguish potholes from deceleration lines. Additionally, when operating, data is sometimes slow and congested due to the use of cheap sensors. “In the near future, the team will use more specialized sensors and provide higher data stability,” Phuc said.

The group wants to coordinate with bus companies in the inner city to automatically collect data on road surface conditions and flooding throughout the city. This data is posted to the interface to monitor and notify promptly to minimize risks for traffic participants quickly and effectively.

According to Associate Professor, Dr. Nguyen Ngoc Lam, former Head of the Institute of Electronics – Informatics and Automation, Ministry of Industry and Trade, in terms of ideas, the group’s solution to warn of potholes and flooding is very practical, helping people. Identify damaged road areas. He said that the group’s pothole warning system is implemented through car shock absorbers equipped with sensors. It should be noted that when the car drives at high speed through small potholes, it can shock very strongly. On the contrary, if you run at low speed and hit a large pothole, the shock will be less. Associate Professor Lam assessed that these cases could affect the accuracy of the level of road damage. “The team’s algorithm currently does not determine the road roughness and damage distance with specific parameters. This is an important requirement to have a basis for accurately assessing the road surface condition,” Associate Professor Lam said.

The group’s product won the consolation prize of the Euréka Scientific Research Student contest organized by the Youth Union of Ho Chi Minh City National University in December 2024.