Data Science Algorithms and Machine Learning Flag Suspicious Tests by Remote or Unsupervised Workers
Impairment Science, Inc. announces the release of a new, breakthrough feature – called Druid True - of its Druid fitness-for-duty app. With this new feature, the Druid impairment test can be taken anywhere, anytime, as opposed to being taken under direct supervision, as previously recommended. Applying machine learning to Druid test data, this new anti-cheat technology flags suspicious Druid test-taking, noting it in the Druid Enterprise online management portal. The portal gives managers the option to have an immediate email or text alert sent to a job supervisor to follow up such suspicion.
“This feature, six months in development, will speed up Druid test taking, already only 1-minute per test, allowing all individuals in an organization’s workforce to take the test on their own time and without direct supervision. This development will also enable organizations with a remote workforce to check fitness-for-duty anywhere, at any time,” said Rob Schiller CEO of Impairment Science. “Already the most sensitive method for measuring cognitive impairment, the new feature creates greater flexibility in testing groups and makes Druid accessible to a much wider range of applications and industries. Ultimately, it will drive improved decision-making regarding fitness-for-duty and help transform managers’ ability to manage on-site risk.”
The Druid® app was first introduced by Impairment Science, Inc. (ISI) in 2016 to assess a person’s level of cognitive and motor impairment due to any cause or combination of causes. The app requires users to perform balance and game-like tasks that measure reaction time, decision-making accuracy, hand-eye coordination, time estimation, and the ability to perform divided-attention tasks. The app has been cited by leading impairment researchers at the Johns Hopkins School of Medicine as “the most sensitive measure of impairment” available. It is being used by academic and corporate researchers both in the US and abroad.
Michael Milburn, PhD, Chief Science Officer, and Max Daniller-Varghese, PhD, Director of Data Science, developed Druid True based on Druid score patterns established by thousands of users. “With Druid, users establish a recognizable configuration for how they take a test that is based on hundreds of measurements,” explained Dr. Milburn. “When a test does not conform to this pattern, based on a normal data imprint, the system raises a warning flag.” Dr. Daniller-Varghese led the project development and added, “Our rich test database provides the foundation for statistical models of test patterns that can be compared to any user and any user’s test history. We expect our model to become even more precise with more field usage.”
Figure 1 - Test Results in Druid Enterprise
With Druid True, tests that show suspected cheating are flagged in Druid Enterprise.
The Druid True feature is available by subscription to Druid Enterprise, Impairment Science, Inc.’s cloud-based management portal, to the benefit of any company in any safety-sensitive industry. For more information, contact Chris Bensley, COO or email info@impairmentscience.com to learn more.