Khaleej Times, Wednesday, Mar 08, 2023 | Sha'ban 16, 1444
How smart driving test system uses AI, sensors to evaluate drivers
Emirates: The Smart Driving Test System (SDTS), which is used to test new
drivers in Abu Dhabi, is an automated driving behaviour assessment that scores a
candidate based on pre-defined testing criteria. This artificial intelligence
system captures every action a driver takes during a test trip.
“As soon as the driver opens the door, the system starts. The cameras record,
starting from seat/ mirror adjustment to speeding, lane changing, leaving safe
distances, compliance with traffic signs, and much more,” said Anas Al Zghoul,
head of research and development at Tatweer MEA.
Tatweer is participating in the Dubai World Police Summit, hosted by the Dubai
Police at the World Trade Center. The summit provides a platform to explore the
current challenges that police forces are facing, as well as the latest
technological advancements, strategies, and innovations across the law
enforcement and security spectrum.
The SDTS car uses state-of-the-art artificial intelligence technology with
advanced sensors and cameras. “We have nine cameras mounted on the car with many
sensors. The module reads the data from the vehicle’s computer- Controller Area
Network (CAN) data. Out of the nine cameras, four are mounted on each side for a
panoramic bird’s eye view and two are mounted on the rear and front windshield
to record the vehicles on the road and driving capabilities of the ride,” said
Zghloul.
“The other two cameras are installed on the front and rear glass panels. The
camera is mounted in front to monitor the driver's eye gaze and eye movement
while changing lanes entering the roundabout, etc. It monitors if the driver
takes the compulsory checks,” added Zghloul.
Tatweer management believes it introduces a new innovative concept of driving
licensing, where automation and process standardization ensure accuracy,
efficiency, evidence, and timely monitoring capabilities. “This guarantees
transparency and reduction of human error through eliminating human
interference,” said Zghloul.
The cameras, sensors, and all the AI module are connected to a panel at the rear
of the vehicle, which relay the data to the centre for evaluation.
Features of the system
- Full Automation
Fully automated process for evaluating driving behaviour based on pre-defined
testing criteria.
- Processing Unit
Centralized data unit that reads data inputs from all system components,
including devices, sensors, vehicle information (OBDII), etc., and then
processes the data through machine learning algorithms to extract the
applicant's testing faults.
- Result Verification
Test findings are verified by examiners and operators to double-check, accept,
or reject.
- ADAS
Advanced Driver Assist Systems (ADAS) is a state-of-the-art technology that uses
image processing algorithms for the recognition of road features to assess the
applicant's reactions and decisions accurately.
- KPIs & Data Analysis
The system produces different reporting of test numbers, test results,
applicants’ characteristics, hardware/ software health status, etc.
- Location Mapping & GIS
The system integrates with the latest GIS technologies to report and monitor the
applicant's exact location in real-time.
- Live Test Monitoring
The system allows supervisors to monitor the tests in the GIS platform and the
vehicle's location.
- Results with Evidence
The system displays the applicant's score, and test result with a Pass/Fail
status supported by video evidence of the applicant's errors for future use and
verification purposes