Smart Transportation Utilizing 3D Scene Reconstruction to Identify Individual Distance and Number

Hardware: reComputer J202 carrier board for Jetson Nano/Xavier NX/TX2 NX module

Use Case Provider: TECHRAIL

Application: People Distance Identification & Crowd Flow Assessment

Industry: Smart Transportation

Deployment Location: Italy

TECHRAIL successfully delivered an innovative solution that enables real-time 3D scene reconstruction. This solution was deployed using the reComputer J202 carrier board for the NVIDIA Jetson Nano/Xavier NX/TX2 NX module. It not only offers insights into the number of people present in the carriage but also provides accurate 3D-mapping distance measurements. Additionally, the system can efficiently transmit real-time results to a central server for effective management, ensuring optimal bus operations at every station.


The initial idea for this concept originated during the pandemic period, a time that undoubtedly represents one of the most challenging and transformative periods in recent history. Beginning in late 2019 and continuing well into 2023, the global pandemic unleashed a wave of uncertainty, disruption, and loss across the world. As a novel coronavirus known as SARS-CoV-2 rapidly spread, governments, healthcare systems, and individuals were thrust into an unprecedented crisis, requiring them to confront the virus head-on. This relentless spread prompted the widespread implementation of lockdowns, travel restrictions, and social distancing measures as means to manage crowds and mitigate the virus’s impact, resulting in the temporary halt of entire nations.


Traditional video monitoring devices that are commonly installed on subways, buses, or school buses have their limitations when it comes to accurately assessing the distance between individuals. These devices can only provide qualitative information such as the number of people present. To overcome this limitation, 3D systems using Time of Flight (ToF) with LIDAR (Time of Flight Laser), LASER, or LED( Time of Flight Camera In Lidar for light detection and ranging) are necessary to conduct a three-dimensional reconstruction of the internal space.

However, in order to determine the distance to each point of the object using LIDAR, the entire object needs to be scanned. This scanning process is typically performed by mechanical actuators, making LIDAR very precise but relatively slow due to the speed of the mechanical actuator’s rotation. In contrast, ToF measures the distance between a sensor and an object by analyzing the time or phase difference between signal emission and reflection. In fact, the ToF cameras provide quick distance estimation for all pixels, but it still has limited spatial resolution and accuracy influenced by ambient lighting conditions.


Techrail has introduced an innovative solution based on passive stereographic technology. Through the development and design of the reComputer J202 carrier board with the NVIDIA Jetson module, Techrail accomplishes data processing and inferencing tasks for real-time 3D scene reconstruction. By combining this technology with object detection models, the system can detect both the distance in 3D-mapping and the number of people in the carriage every 2 seconds. Typically, one 16m bus can be thoroughly covered by 3 Right Metro boxes. The entire system provides local data storage and transmits the real-time results to a central server. Control staff can manage information on each individual carriage and transfer it to information panels when the train arrives at each station.

After being deployed and thoroughly tested in Milan, it has been discovered that the margin of identifying accuracy error can be tightly controlled within a mere 1%. Each sensor is capable of providing instantaneous analysis of the prevailing situation. Furthermore, the grids are specifically designed to indicate the exact position of passengers who have returned to the area that is visible by the corresponding sensor. In cases where at least a mutual distance from other passengers is less than one meter, a red cross is prominently displayed.


TECHRAIL is an innovative start-up of researchers, scientific research centers, universities, and the business community that draws on the experience of professionals with more than two decades of experience in the innovation, design, and development of transport technology systems. Defense, telecommunications, and industry. They are a flexible and “market-driven” organization that puts skills in the field of research at the disposal of companies to generate original and innovative solutions aimed at continuously improving the operational performance of our customers through technology transfer.

Seeed NVIDIA Jetson Ecosystem

Seeed is an Elite partner for edge AI in the NVIDIA Partner Network. Explore more carrier boards, full system devices, customization services, use cases, and developer tools on Seeed’s NVIDIA Jetson ecosystem page.

Join the forefront of AI innovation with us! Harness the power of cutting-edge hardware and technology to revolutionize the deployment of machine learning in the real world across industries. Be a part of our mission to provide developers and enterprises with the best ML solutions available. Check out our successful case study catalog to discover more edge AI possibilities!

Take the first step and send us an email at [email protected] to become a part of this exciting journey! 

Download our latest Jetson Catalog to find one option that suits you well. If you can’t find the off-the-shelf Jetson hardware solution for your needs, please check out our customization services, and submit a new product inquiry to us at [email protected] for evaluation.

About Author


June 2023