NSF-Funded Student Design Projects:
NSF-Funded Student Design Projects:
ETL Matching Correspondence Counter
Designers: MultiSemester see Student Teams Table
Client Coordinator:Donna Case, Ph.D., OTR, Northville Public Schools
Supervisors: Dr. Robert Erlandson, Mr. David Sant
Department of Electrical and Computer Engineering
Wayne State University
Detroit, MI 48202
Multi-Semester Project Student Teams
| Team Members | Date | Project Component |
| Ray Ryan | Fall 2001 | Researched sensor types.Basic sensory cell design & prototype. |
| Craig ElderShahiar Kamal | Winter 2002 | Sensor Cell & System Packaging. |
| Kristine PattersonGuran Jancevski | Winter 2003 | Microcontroller. |
| Fan Yang | Winter 2003 | Final system integration. |
Introduction
Counting is a problem for students with cognitive disabilities. Teachers and therapists often use a matching correspondence approach to counting. For example, if ten items need to be counted for packaging, the job coach would lay a piece of cardboard on the worktable which contained a 2x5 grid of squares. The student would then be instructed to place one item in each square until all the squares were filled, hence the term matching correspondence. When the student has filled all the squares the job coach removes the items for packaging.
The matching correspondence counter (MCC) grid varies depending on the job and the ability of the student. The grind can vary from 1x2 up to 5x5. However, as the number of cells in the counting grid increases the matching correspondence counting task gets more difficult for the cognitively disabled students. The students tend to place objects on the counting grid randomly. As the number of grids increases the students slow down. They seem to have problems figuring out which grid cell to use for placement. Recent work suggests that the student’s slow down is an expression of Hick’s Law. The students are presented with a large number of competing decisions and as the number of choices increase it takes longer to make a decision.
The Matching Correspondence Counter (MCC) is designed to support students with counting and packaging tasks. The ETL Coordinator enters into a dialogue with the MCC and student and guides the student through the counting task. The MCC is one of a family of “smart sensors” used with the ETL Coordinator. A teacher or job coach uses the ETL Coordinator to configure the task requirements; number of items to be counted, prompt times, and prompting messages.
Students are prompted by the Coordinator to place an item into a cell where the indicator lamp is turned on. When an object is placed into the cell, the indicator lamp is turned off and the student prompted to continue fill cells whose indicator lamps are lit. When the required count is reached the Coordinator informs the student to stop and package the counted items. If the required count is more than ten, the number of cells in the MCC, the Coordinator keeps track of the total count and informs the student to empty the cells into the packaging container. As the MCC cells are emptied, the Coordinator turns the indicator lamp back on. When the specified count has been reached the student receives the job done prompt to empty the cells into the packaging container.
Summary of Impact
A prototype matching correspondence counting device has been designed and assembled. Figure 1 shows the completed device. There are ten instrumented counting cells. The device has not yet been field-tested. The field-testing is planned for the 2003-2003 academic year.
Figure 1. The prototype matching correspondence counting device.
There are ten instrumented cells for the counting process.
The device interfaces with the ETL Coordinator. The Coordinator provides the logic for
determining when a counting goal has been successfully reached.
Technical Description
Figure 2 shows the instrumented counting cell. Each cell has a small indicator lamp. There is a row of infra-red emitters and detectors at the bottom of the cell which detect the presence of an object in the cell. The inclined plane is made of anti-static plastic.
Figure 2. Left - a single cell. Right - the internal sensing structure.
Students drop an object into the cell. The anti-static inclined plane guides the object to the bottom of the cell where it will break the IR beam between at least one emitter – detector pair. The emitters and detectors are connected via an “OR” logic so that any break will trigger the presence of an object.
The MCC utilizes a Microchip PIC16F876 flash microcontroller to monitor each cell and control the illumination of the LED prompting indicator. The PIC16F876 microcontroller also handles communication between the sensor bins and the ETL Coordinator System. A block diagram of the MCC is shown in Figure 3.
Figure 3. Block diagram of MCC
The MCC’s microcontroller periodically scans the bins and filters out any electrical noise that may be introduced in to the system. The results of the scan are stored in the controller and are accessed over an RS-232 serial link by the ETL Coordinator. The Coordinator System has complete control over the MCC and using simple ASCII based commands can perform the following functions:
1) Get the MCC’s Firmware Version (Useful for attached device identification)
2) Turn On the MCC’s LED emitter power supply
3) Turn Off the MCC LED emitter power supply
4) Set the prompting Indicator LED
5) Get the bin occupancy status
6) Get the number of connected bins (Useful for diagnostic purposes and configuration)
A typical communication flow begins with the coordinator inquiring about the attached devices firmware. From this information it knows that it is connected to a MCC unit. From there the coordinator issues a command to turn on the LED emitters. After a brief stabilization period the coordinator performs a self test of the prompting LEDs to verify that they are working. The Coordinator then determines how many bins are connected to the system. Once this information is retrieved the Coordinator system is ready to be used.
During use the coordinator periodically queries the MCC for bin status, issues user voice prompts, and turns on visual indicators based on the input received.