Loss of Privacy

The Huntsville school district aims to keep track of students who ride the bus with a new RFID system.

the system, called ZPass, will allow school administrators to keep better track of those students who ride the bus each day. Kyle Koski, transportation director of the district, said about 5,000 students ride the bus.

“It is an immediate way that we can have feedback if a child does not get off (the bus) where he’s supposed to,” Ward said.

The district and police went on alert twice in August when, on the first and second days of school, two elementary students briefly went missing after taking the wrong buses.

This system is being implemented because two small children got on the wrong bus on the first and second days of school. This sort of thing happens every year when students try to learn the new bus system. A student may have ridden bus #7 last year, but this year bus #12 runs the route that takes him/her home. The same applies to new students who don’t know the bus system, students who have never ridden the bus before and kindergarteners who simply don’t understand how things work.

Each student in the pilot program will be assigned a personalized radio frequency identification (RFID) card, which they will swipe in front of a card reader installed on the bus’ dashboard. Students will swipe their cards each time they get on the bus and whenever they exit.

Using RFID technology and GPS, the card reader records the location of the bus at the time of the swipe and immediately loads that information onto the district’s computer network. At any time, administrators can pull up the data — including a map — and see exactly where a student both entered and exited a bus.

Though the new system is designed to track students, it has a major flaw. The system appears to only match a head count with the count on the RFID reader. There’s nothing to say the person holding little Johnny’s card is really Johnny. This is not a new flaw, yet every district that attempts to use RFID for tracking students seems to ignore it.

photo by Mark W. Stroud

In a California military base, marines are being trained on how to collect biometrics from civilians. Intended for use in Afghanistan, this training could have implications elsewhere.

Military policeman with Combat Logistics Battalion 4 participated in tactical site exploitation training at the training city of Wardah-Mir, Marine Air Ground Combat Center Twentynine Palms, Calif., Nov. 18.

The Marines conducted foot-patrols, room clearing and search operations where they collected biometric data and other evidence on citizens displaying suspicious behavior or possessing contraband.

“We are trying to give the Marines the skill set to assist the Afghan government in criminal prosecutions and to help teach the Afghan National Police these skills,” said Patrick Garrahan, law enforcement professional, Tactical Training Exercise Control Group, MAGCC Twentynine Palms.

Suspects were processed using a handheld interagency identity detection equipment system, which compared their biometric data, such as fingerprints and iris scans, to a central database, according to Apsey.

The training is meant as a way to help marines detect those who make bombs though how biometric data will do that is unclear. One must first collect biometric data on individuals that, presumably, have done nothing wrong. Then, later, when a person is suspected of making a bomb, their identities can be verified with the biometric data already in the military database.

The FBI is adding facial scanning, iris scanning, and palm scanning to its biometrics databases at the Criminal Justice Information Services Division (CJIS) and its getting some help from the DOD to accomplish this mission.

CJIS is responsible for information repositories–such as the National Crime Information Center, the Interstate Identification Index, and the National Instant Criminal Background Check System–that provide law-enforcement officers with real-time data on people’s criminal history, stolen property, missing persons, and other information.

The additional biometric information will be added to the system that can already track fingerprints.

CJIS processes about 140,000 requests a day through the system, double the number it could handle on a good day a few years ago, he said. Moreover, the algorithm is allowing the FBI to match fingerprints at 99% accuracy versus 92%, which was the previous norm.

The FBI also added facial-recognition and iris-scan systems to its biometrics matching system–which is gradually replacing its predecessor, the Integrated Automated Fingerprint Identification System–and next year will be able to match palm prints for the first time, he said.

To further its work in biometrics, the FBI is teaming with the Department of Defense to build a Biometrics Technology Center on its FBI campus in Clarksburg, West Virginia, Cutherbertson said. The center, which will focus on research to advance biometrics technology, is due to be completed in spring of 2014. “It will be a tremendous resource to carry us into the future,” Cutherbertson said.

With the inclusion of multiple points of biometric data, the FBI hopes to improve security at home. The research conducted at the new Biometrics Technology Center will also allow the FBI to accomplish one of its other goals, biometrically identifying individuals on the internet.

Shopper Tracker does just what it says it does, track shoppers as they move throughout the grocery store. The new twist in this tracker is that it uses a Microsoft Kinect to do so.

It analyzes customer movements to provide traffic flow analysis and heat maps indicating which shelves are attracting shoppers and which products they touch or take. This can be tied to conversion data by product SKU to help merchants optimize where products are placed within their stores.

To attain similar market research, merchants typically have to pay observers or use equipment and surveys that are expensive, inaccurate, and influence the behavior of the people they’re studying. With Shopper Tracker, multiple shoppers can be simultaneously tracked around the clock. As it doesn’t cost much more to conduct longer studies, merchants can get more confident results and do A/B testing.

The data can then be used by the store to determine which items are more popular and why customers go to a particular item, but never purchase it. The system is also capable of tracking multiple people simultaneously.

Shopper Tracker: tracking real world conversions like web analytics from Administrator Agile Route on Vimeo.

All the following metrics and reports can be filtered for any date/time range.

Visitors: people that are detected in the tracking zone. Associated metrics:
Visits and Visit duration

Zones: places where Visitors may physically be (e.g.: right in front of a particular part of the shelf). Associated metrics:
% Visitors through Zone and Average time in Zone

Area: a 3 dimensional space containing products the Visitor may touch (e.g.: an imaginary box right in front of one particular SKU in the shelf)
Event: an Event occurs when a Visitor touches an Area in the shelf. Associated metrics:
Total Events and Events per Area
Average Event time
Visits with Event and Events / Visit Ratio

Conversion: grabbing a product from an Area (requires image captures post-processing). Associated metrics:
Conversions and Conversions per Area
Average Conversion time
Conversions / Events
Visits with Conversion and Conversion / Visit

Goals: grabbing a SKU of interest (e.g.: own product SKU). Associated metrics:
Goal Funnels
Product traction

Other type of reports: Heat Maps
Available reports:
- Visitor transit paths
- Visitor touch location in the shelf
- Conversion locations in the shelf