Voice Biometrics Goes Back to Jail

In an earlier posting, I described the use of voice biometrics in jails/prisons to prevent illegal activity by inmates. An article in today’s Mt. Vernon Register-News reinforces the need for voice biometrics to prevent inmates from perpetrating phone scams.

According to the item, inmates at the Jefferson County Detention Center (Illinois) are using pay phones to call unsuspecting citizens and scam them. “This time, they are using a little different technique,” Jefferson County Sheriff Mulch explained. “The caller uses a calling card that has the ability to program it to Spanish. The person receiving the call only understands the ‘Jefferson County Detention Center,’ portion, and they hit the prompt to speak with the person. They get the detainee, who then speaks in English.”

The detainee pretends to be a police officer who is trying to identify someone who has been involved in a traffic accident, and the phone number they called was on the victim’s cell phone.

“People want to help and ask what they need,” Mulch said. “They are told to dial star 72 and enter a series of eight numbers, or they tell them to dial 1172 and a series of numbers. What that does is locks the phone number in and they can then use a calling card to make calls across the country using the victim’s phone number.”

Sheriff Mulch is in the process of researching the use of voice biometrics to track inmates’ fraudulent calls.

American Safety Council uses Voice Biometrics to Validate Identity of eLearning Students

The American Safety Council, an Orlando Florida based provider of eLearning courses, uses voice biometrics to validate student identity for defensive driving programs in New York and New Jersey. At various intervals during the course, students are required to validate their identity, via voice biometrics, within 60 seconds.

Two months ago, I posted an item about a similar voice biometrics application being offered by I DRIVE SAFELY. Does anyone know if they one and the same?

Alternative to Voice Biometrics? Online "Handwriting and Gesture" Biometrics

ThirdFactor.com recently wrote about a technology company located in Dallas, Texas, Biometric Signature ID, which developed an online handwriting and gesture biometrics application -- BioSig-ID™. According to the Company’s Website, BioSig-ID is an ideal solution for verifying the identity of people using online applications (e.g., distance learning, online banking, etc.). The BioSig-ID enrollment and verification processes work as follows:

Enrollment Process
• An individual is asked to draw their unique signature with any characters/shapes/initials, using their mouse, in an online box. This process is repeated three times to create a “base” signature
• The individual is then directed to a second non-biometric layer of security called Click-ID Technology (i.e., they click on a selection of objects) to complete the enrollment process. There is even a third-level of security using complex security questions

Verification Process
• An individual accesses the online application and is asked to draw their signature once. This process creates a “sample” signature
• If the individual’s “sample” signature matches their “base” signature, access is granted to the application
• If the signatures don’t match, additional levels of security are activated

BioSig-ID was recently selected to complete a pilot trial at University of Maryland University College.

Voice Biometrics Facilitates Automated Password Resets

At last week’s Voice Biometrics Conference, Mike Matthews, UK Service Desk Product Manager at Atos Origin (IT services company), presented a compelling case study on the use of voice biometrics to automate password resets.

Mike started off by listing some examples of password reset “pain” compiled from Atos Origin clients:
• Up to 25% of all contacts to the help desk
• Up to 15 minutes to reset a password
• Up to $22.50 per contact
• Post weekend and holiday spike in contacts
• Support unavailable outside of normal office hours

He then proceeded to describe a fully-automated approach to password resets, using voice biometrics to verify the user’s identity, and identified the benefits. Using an automated approach…

• Resetting a password takes, on average, 1 minute
• The need for human intervention to complete low level tasks is reduced, significantly lowering overhead
• Password resets are not affected by heavy call volumes and, therefore, spikes in contacts become a thing of the past
• Password reset is available 24/7 365 from anywhere in the world, using any type of phone

Mike concluded with the following ROI example:

Company performs approximately 115,000 password resets per year:
• Currently costs $12 per call, or $1,380,000 per year
• Lost productivity, using an average of 7 minutes per call, amounts to 13,500 hours

Using an automated password reset (with voice biometrics) approach, the cost per contact is reduced to $7.50 per call, or $862,500 per year:
• $517,500 savings per year
• Help desk downtime reduced to approximately 2000 hours (i.e., reduction of 11,500 hours per year)

Voice Biometrics Case Study - T-Mobile Netherlands

At last week’s Voice Biometrics Conference, Daniel Hendling, International Program Manager, Sales & Customer Service at Deutsche Telekom, gave a lively presentation of T-Mobile Netherlands’ plans for its upcoming voice biometrics pilot. According to Daniel, complex methods of customer authentication, currently in place, can harm both customers and businesses. For example, customers may get annoyed having to remember different passwords and businesses may see an increase in call handling time – which can lead to customer satisfaction and loyalty issues.

T-Mobile Netherlands selected Agnito’s KIVOX voice biometrics product to achieve the following objectives:
1) Reduce the average handling time spent on caller verification
2) Provide an enhanced user experience
3) Enable reliable self service options in the short to mid-term

Enrollment
• The customer will be prompted to provide their PIN code and repeat three sentences, three times (i.e., “base” voiceprint will be captured and stored in the database)

Verification
• The customer will be prompted to repeat one sentence once (i.e., “sample” voiceprint is captured)
• If the “sample” voiceprint matches the “base” voiceprint, the customer’s identity is verified

T-Mobile Netherlands expects to realize the following benefits from their voice biometrics application:
• Improved customer satisfaction
• Decreased call handling time
• Increased level of security
• T-Mobile Netherlands will be perceived as an innovation leader

The pilot is scheduled to commence July 2010. Stay tuned!

Voice Biometrics to Monitor Florida Home Health Providers

Sandata, LLC, a provider of IT solutions to the home health care industry, has announced the upcoming deployment of a voice biometrics system that will enable the Florida Medicaid program to verify that home health services were actually delivered to the recipient. The pilot, scheduled to begin July 1, 2010, will target home health providers in Miami-Dade County that participate in the Florida Medicaid program. According to Harold S. Blue, Sandata’s CEO, the voice biometrics system will help ensure that the "The Florida Agency for Health Care Administration (which administers Medicaid) only pays for the '5 rights' - the right service provided, at the right location, at the right time, for the right patient, by the right caregiver."

The voice biometrics system, which is powered by PerSay’s VocalPassword™ product, will confirm home health visits by taking calls from home health personnel (nurses and home health aides), who directly provide services to Medicaid recipients, at the beginning and end of each home health visit using the recipient's home telephone. The system will verify that the correct caregiver provided the service by comparing their sample voiceprint to their base voiceprint on file.

Voice Biometrics Conference 2010

I just returned from the annual Voice Biometrics Conference – where the key players of the voice biometrics industry gathered to compare notes, present case studies and discuss future plans. Kudos to the Opus Research team -- Dan Miller, Derek Top and Pete Headrick -- for orchestrating an excellent event.

Over the next week, or so, I plan to share some new case studies that should resolve any lingering concerns about the efficacy of voice biometrics technology. The bottom line is that voice biometrics systems have been deployed at leading organizations throughout the world, and are successfully identifying and verifying individuals by their voice 24/7.

My take away from the conference is as follows: the voice biometrics industry offers an excellent solution, with a compelling value proposition, but still appears to be treading water in the market place.

New iPhone biometrics application? Not sure…

An article on “The New Zealand Herald” Website, states that a New Zealand technology company, FeelGudApps, has developed a new biometrics application for the iPhone. The developers claim that the app, named ‘Feelie’, can broadcast how you're feeling to registered recipients. The item goes on to describe how Feelie works.

Feelie monitors a person’s behavior (e.g., terseness of texts and emails, amount of movement, etc.) for two to three days. A base profile is then constructed based on data from psychological studies. From then on, the person's emotional state is dynamically updated and broadcast to their registered recipients. A color-coded tinge over the person’s profile picture identifies how they're feeling as follows:

• Yellow – happy
• Blue - calm
• Green – nervous
• Purple - lying
• Red – angry

I was hoping to put Feelie through its paces -- to see if there was a voice biometrics angle to be exploited. Unfortunately, I could not find any reference to the app in the iTunes store. Too good to be true?