Over the past months, The DADSS Research Program and our partners at Senseair™ have been collecting data from in-vehicle tests (featuring Generation 3.1 Breath-based Sensors) through the program’s Pilot Field Operational Trial and Driven to Protect Partnership with Virginia DMV and James River Transportation. We have used the data from these on-road tests to significantly improve the alcohol measurement for the next generation sensor, GEN 3.2.
For example:
- We have improved the airflow through the system and have also improved the alcohol detector, giving us greater sensitivity to naturally exhaled breaths.
- We have enhanced our measurement algorithms so they can analyze several consecutive signals, providing us more data for more reliable measurement.
- We have identified automotive mass production-friendly options for key components of the technology—such as the detectors, emitters, and mirrors—that will help maintain quality while enhancing manufacturability.
- We have upgraded the sensor calibration procedures to be used at the point of manufacture to be consistent with those used at our verification and validation lab to ensure calibration accuracy and reproducibility.
- We also have put the GEN 3 sensors through rigorous environmental testing that simulates a sensor lifetime of fifteen years.
As a result of all these advances, the GEN 3.2 breath sensors now are sufficiently advanced for on-the-road testing.
On the Road
In September 2018, the DADSS Program joined the Virginia DMV to publicly announce Driven to Protect and its partnership with James River Transportation (JRT) to conduct in-vehicle tests on Virginia roads with JRT drivers and fleet vehicles. These vehicles—equipped with GEN 3.1 sensors—have logged over 3,800 hours, obtained over 18,000 samples and driven 40,000 miles to date, and have provided invaluable information to help improve our algorithms.
We are also conducting on-road testing in Massachusetts. Our engineers have logged 5,000 hours, obtained over 33,000 samples and driven 24,000 miles to date with the GEN 3.1 sensors to make improvements to vehicle integration and to improve the ruggedness of our data acquisition systems. We are now in the process of equipping 10 additional vehicles with the GEN 3.2 sensors and look forward to seeing how they perform in the New England winter.
In the Labs
We have recently received additional GEN 3.2 sensors for internal testing. In our labs, we assess the performance of each sensor before integrating it into a vehicle and each sensor will be tested again after it has been on the road—tracking the changes to determine potential improvements to the technology. We have been able to dramatically increase the speed with which we can conduct this battery of tests: in 2018, we have been able to test eight at a time and in the first quarter of 2019, we expect to be able to test 24 at a time.
We have also moved the creation and mixing of the gas we use to analyze our sensors in house—allowing us to exceed our performance specifications by a factor of five and eliminate additional variables in our testing. We are also working on a portable “breath sample” system.
We continue to test and evaluate the sensors using human subjects in a controlled laboratory setting at McLean Hospital, a Harvard Medical School affiliate. The DADSS Program is currently verifying GEN 3.1 sensors and will begin incorporating GEN 3.2 into the next round of testing. Once GEN 3.2 is performing as well as or better than GEN 3.1, GEN 3.1 will be eliminated from testing.
Voice Activation
To demonstrate to partners and the public at large just how passive the GEN 3.2 sensor has become, the DADSS Program developed and patented a voice activation system. By speaking a sentence of a specific length (“I am driven to protect, verify me”), the driver produces enough of a breath sample for the sensor to provide an accurate breath alcohol reading. You can view Governor Northam testing this system at the Driven to Protect event here.