FMCG

Safeguarding Commuter Safety: Real-Time Bus Monitoring for Fortune 500 Company

<1 Second

Reaction Time of the System

95%+

Accuracy in Detection

INTRODUCTION

In an era where employee well-being is at the forefront of corporate responsibility, a leading multinational company provided shuttle bus services for its staff, ensuring a safe commute between their homes and offices. However, a concerning pattern emerged. Some passengers, in their hurry, would occasionally disembark while the bus was still in motion, leading to accidents. Recognizing the gravity of this issue, the company sought an innovative solution to ensure the safety of its employees during their commute. They presented Codemonk with this challenge, and we got right to work in figuring out a solution

Industry

FMCG

Time Frame

6 Months

Services we offered

Computer Vision, ML on Edge, IOT

https://content.codemonk.inGroup 21677.webp

The Challenges

Real-time Monitoring

The heart of the issue was the need for vigilant, real-time surveillance of the bus's interior. It was crucial to promptly detect any movement or activity suggesting a passenger's intent to disembark while the bus was still in motion. This real-time detection was pivotal to prevent potential accidents and ensure passenger safety.

Instant Alarm & Notification System

The challenge was to devise a system that could instantly sound an alarm in the bus if a passenger moved in the aisle. Additionally, this system needed to simultaneously generate a security report, including the incident's video, and send it to the office team.

Offline Functionality

Buses often traverse areas with patchy or no internet connectivity. This posed a significant challenge as the solution had to be self-sufficient and capable of operating offline, ensuring that monitoring remained consistent and uninterrupted throughout the journey, regardless of the connectivity scenario.

Data Storage and Synchronization

While offline functionality was essential, the system also needed to store and, when possible, synchronize data with a central database.

Codemonk’s Approach

After in-depth discussions with the client and researching and analyzing the buses, our team envisioned an integrated solution that seamlessly combined both hardware and software elements. We strategically chose a small, credit-card sized computer paired with night vision cameras, ensuring adaptability to fluctuating lighting conditions and a compact footprint for hassle-free bus installations. But the real innovation lay in our use of ML on edge. By deploying AI directly on these compact boards, we enabled real-time movement detection without the need for extensive computational resources or constant cloud connectivity. This setup, complemented by our custom software algorithms, ensured immediate and accurate detection of any movement within the bus, making our solution both smart and efficient.

The Solution

Deployment of Real-Time Monitoring System

We strategically deployed a compact package within the buses, which included Raspberry Pi boards, night vision cameras, accelerometers, and other essential sensors. This comprehensive setup was designed to detect any movement in the aisle, even in varied lighting conditions. The compact nature of our solution ensured easy installation, minimizing disruptions and making it adaptable to different bus interiors.

ML on Edge for Real-time Detection

Harnessing the power of Machine Learning directly on the Raspberry Pi boards, we enabled instantaneous movement detection. This "ML on edge" approach ensured that the system could process data in real-time, without the need for cloud connectivity, making the detection process faster and more reliable.

Instant Alarm System

Upon detecting any movement, our system was designed to trigger an immediate audible alarm within the bus. This alarm served as an immediate deterrent, alerting the passenger and the driver about the potential safety risk.

Security Reporting with Video Evidence

Simultaneously, when movement was detected, our system captured video evidence of the incident. This footage, along with a detailed incident report, was then transmitted to the security personnel at the client's headquarters, ensuring they were always in the loop about any potential safety concerns.

Synchronized Speed Monitoring

Our system integrated sensors that simultaneously monitored the speed of both the bus and the detected passenger movement. Tracking movement within a moving vehicle presents unique challenges, and by gauging the pace of both the bus and the passenger, our solution ensured accurate and instant detection, reducing false alarms and enhancing safety measures.

Impact

The implementation of our solution brought about a transformative change in the safety protocols of the company's shuttle bus services. Here's a snapshot of the tangible impact:

Enhanced Safety Measures

The real-time monitoring and instant alarm system drastically reduced instances of passengers attempting to disembark prematurely. The immediate audible alarm acted as a deterrent, ensuring passengers remained seated until the bus came to a complete stop.

Operational Efficiency

By automating the surveillance process, the need for manual monitoring was eliminated. This not only reduced the operational overhead but also ensured consistent and reliable monitoring across all shuttle buses.

Positive Employee Feedback

The employees, who were the primary beneficiaries of this solution, expressed increased confidence in the shuttle bus services. The knowledge that their safety was being actively monitored and prioritized led to higher satisfaction levels and trust in the company's employee welfare initiatives.

The Final Word

Our collaboration with the client was not just about deploying a tech solution; it was about ensuring the safety and well-being of thousands of employees. The success of this project underscored the importance of innovation in addressing real-world challenges. We're proud to have played a pivotal role in enhancing the safety standards of a global corporate commute system.

<1 Second

Reaction Time of the System

95%+

Accuracy in Detection