Driver Impersonation
The aim of this feature is to prevent driver impersonation by implementing facial recognition technology for DPs.
Introduction
In the rapidly evolving ride-hailing industry, ensuring driver authenticity is crucial for maintaining trust and safety. Driver impersonation poses a significant risk, potentially compromising passenger security and platform integrity. To address this challenge, we implemented a facial recognition system that verifies the identity of driver partners (DPs) in real time.
By seamlessly integrating this verification process into the driver app, we aim to mitigate impersonation risks, uphold compliance standards, and enhance overall platform security. Additionally, this solution will streamline the support process for DPs who may be mistakenly flagged for impersonation.
This case study explores the research, design considerations, and implementation of the facial recognition feature, ensuring a balance between security, user experience, and operational efficiency.
Challenge
Current Behavior
Driver partners ask someone else to work on their behalf using their account. This can be a person who was not trained and verified by us.
Competitor Analysis
Multiple delivery based companies and ride hailing companies consists of feature which leads to identity verification, Thus taking inspiration from some of them and built a better for BluSmart in order to mitigate the risk of driver impersonation.
Multiple Iterations
Iterations were based on following points:
As a DP, I want to capture an image on my app for verification to prove my identity.
As a DP, I want a simple process for capturing and submitting the verification image.
As a DP, I expect immediate feedback on the success or failure of the image verification process.
As a DP, I want an option to retake the image if it doesn't meet verification criteria.
As a DP, I want the ability to contact the support team if my account is blocked due to impersonation.
Feature Solutioning
Driver App:
Picture Capturing Process
During the first login after the feature launch, DPs will be required to capture an initial image for verification.
All captured images will be securely stored in the database for identity matching.
Camera Interface & Face Detection
A circular frame will be displayed on the screen to guide face alignment.
The circle will remain red until a human face is detected.
Picture Matching & Verification
After capturing the image, DPs will have the option to review the picture before submission.
If the captured image is clear and satisfactory, the DP can submit it for verification.
If unsatisfied, the DP can retake the picture and repeat the process.
Hub Dashboard:
Impersonation Case Management
A new 'Driver Impersonation' section added under the Drivers Tab in the Hub Dashboard.
This section will store and display all impersonation-related cases for quick access by the OPS team.
Metrics Achieved
DP Impersonation feature thus reducing the frauds of impersonation and bring it to nearly 0 week on week.
