Hitachi develops breath-alcohol detection device with facial recognition
Hitachi has developed a prototype system for a new breath-alcohol detection device with facial recognition capabilities.
Since developing a prototype with Honda R&D Co. in March 2016, Hitachi has been conducting further research to increase the reliability of the sensor, reduce its size, and develop application software.
In March 2017, Hitachi conducted field tests involving employees at three sales offices of Hitachi Capital Auto Lease to verify the practicality of the new breath-alcohol detection device — namely, to figure out whether the mobile prototype can collect and manage drivers’ alcohol test data on smartphones.
The prototype provided the convenience of testing the device in any location, and prevented tampering by being able to distinguish whether or not the sample submitted was human breath.
However, the team identified two key issues: first, the possibility of abuse by a substitute taking the test instead of the actual driver; and second, the need to be able to centrally manage test data from each driver in fleet situations.
To address these issues, Hitachi developed a function to prevent the use of substitutes as well as enabling central management of alcohol test data.
Hitachi reduced the size of the new prototype detection device to about one-third that of the previous prototype by eliminating the battery and related circuits, while maintaining the same accuracy for alcohol detection.
As a result, users are able to attach the prototype to a smartphone, and then initiate the software to take the alcohol breath test. During this time, a facial image of the individual is simultaneously captured.
Following the test, the smartphone is placed in a holder installed in front of the driver’s seat and another photograph of the driver is captured.
The system then matches the image of the person who took the test with that of the driver to confirm identity as well as prevent substitutes from taking the tests.
The team also developed an application to send the breath-based alcohol test results from the detection devices to smartphones and aggregate the data.
Safety administrators can download the log data related to aggregated time of test, presence or absence of alcohol and mobile terminal ID to a smartphone or PC, which increases management efficiency and enables remote management of alcohol inspection.
Hitachi is expected to introduce the breath-alcohol detection device with facial recognition capabilities as early as August.
Earlier this year, Hitachi developed detection and tracking technology using artificial intelligence (AI) which can distinguish a person in real-time using features from over 100 categories of external characteristics including sex, color of clothing or items in hand.