IMAGE CAPTURE:Smart kiosk helps consumers recycle cell phones

Every other month, it seems, cell phone manufacturers introduce new products that are more compact, offer more applications, and longer battery life than previous generations. Eager to purchase these products, many consumers now find that they have one or more obsolete cell phones in their homes or offices. While many web-based companies now allow consumers to mail in these products in exchange for cash, ecoATM (San Diego, CA, USA) has developed a more convenient method in the form of an automated recycling station.

“In the design of the eCycling Station, it was necessary to make the task of recycling as easy as possible for the consumer, and be able to pay them on the spot,” says Michael Librizzi, vice president of business development and co-founder of ecoATM.

In operation, the station prompts users through a touch-screen interface to place their cell phone into an inspection bin at the front of the machine. Before placing the cell phone in the inspection bin, however, a rectangular area within the bin is illuminated with a projected image to show the user where to place the device (see figure, part a).

a) Consumers can now recycle their cell phones at eCycling Stations from ecoATM by using a touch panel interface to place phones into an inspection bin at the front of the machine. b) The cell phone is then analyzed visually and electronically and a receipt for the value of the phone dispensed.

After placing the cell phone in this location, the bin is then closed and illuminated in both on- and off-axis configurations using banks of red LED panels placed on all sides of the bin. In this way, the optimal contrast can be achieved independent of which type of telephone has been placed in the bin.

After capturing monochrome images of the telephone with a scout camera from Basler (Ahrensburg, Germany; www.baslerweb.com), the data are transferred over a FireWire interface to a host computer, where data from the image are preprocessed, and features are extracted and processed by the PC Cure Pattern ID Toolkit Virtual Instrument Library (VIL) from Neural ID (San Mateo, CA, USA; neuralid.com).

Neural network-based software allows extracted features to be rapidly compared with features from images of the more than 4000 different varieties of cell phones stored in the system’s database. In this way, the type of telephone and its interface can be made.

In addition to determining which type of telephone is present, the system also analyzes the image of the phone for any damage such as scratches, nicks, or missing keys and grades its appearance in eight different classes.

To allow the user to erase any stored data, the eCycling Station must present the consumer with a cable to connect to the telephone. “In the over 4000 different types of telephones that the system can process, more than 95% of valuable phones use just 20 different connector types,” says Librizzi. To present the user with the correct type, 20 different cables are arranged across an x-y positioning system equipped with a motorized actuator.

After the system determines which telephone is present and which connector must be used, data are transferred from the PC that moves the x-y positioning system to the correct position. The motorized actuator then plunges the cable into the inspection bin for the consumer to connect to the telephone. The system then performs an electrical test on the telephone and presents the user with an option to sell the telephone at a specific price.

To determine the value of the phone, a second networked PC stores pricing data based on the type and quality of the unit. The data are updated from a central database by potential purchasers of the phone. After accepting the price offered by the eCycling Station, any data are erased and the cable is withdrawn from the inspection bin, effectively disconnecting the telephone. The bin is then rotated and the cell phone stored in the recycling station for later retrieval. A video showing how the user interacts with the station can be seen at http://bit.ly/b18Ooh.