Dehydration Monitor

Team 2Kevan Chao

Brandon ClarkAlexandra Deal

Ian GunnSamaria Mitchell

Dan Schlegel

Presentation Outline

• Problem Identification• Customer Requirements and Tech. Specs.• Alternative System Concepts• CAD Diagrams• Subsystem Breakdown• Validity of Tests• Future Design Modifications

Customer Requirements and Performance Specifications

Customer Requirement #

Technical Specifications Target Values/ Range of Values Actual Value/ Range of Values

1 Accuracy and Precision 99% reproducibilityAccurate within 3% of hydration levels

99.213% accuracy

2 Generated force by pressure device

<2 lbs. <2 lbs.

3 Amount of chemicals or hazardous waste

0 lbs. of hazardous waste and chemicals 0 lbs. of hazardous waste and chemicals

4 Blood drawn 0 cc 0 cc

5 Number of buttons 1 – start button 1 –start button

6 Time to complete testing < 10 seconds < 10 seconds

7 Font size Size 24 on laptop screen Size 24 on laptop screen

8 Time can be exposed to certain weather conditions before damage to system

Lasts over 5 minutes Lasts over 5 minutes

9 Low power consumption 8 Watts 5 Watts

10 Size Excluding laptop fits within 9” x 9” x9” 3.5” x 4” x 4.5” box, excluding laptop

Problem Identification

• Dehydration– Causes effects such as nausea, headaches, increased respiratory

rate, and weakness– Measured as a percentage (weight of fluid lost)/(maximum fluid

weight in body)– Major concern for endurance athletes and students in ROTC –

2% dehydration decreases performance– Concern for all persons – 5% dehydration requires

hospitalization• Market Competition– No current product on the market accurately displays

dehydration levels based on actual user levels

Alternative System Concepts

• Sow Machine Concept– Bulky– Unsafe– Needed to design a

servo system to push dowel

• Pros– Inexpensive

CAD etc.

Final Concept

• Box Design Concept– Unaesthetic

• Pros– Simple[servo]– Relatively

Inexpensive– Durable

CAD etc.

System Function Overview

Start code• Call Arduino from

Matlab• Call Matlab function

Test Begins• Base Image Captured• Pressure applied by

servo• Return servo to

starting position

Comparing Images

• Capture sequence of images

• Pixel-by-Pixel comparison

• Compare to threshold (25%)

Output• Dehydration level

based on frame campture

• Consult chart to rehydrate

Pressure Delivery

• Original plan was for dowel system• Changed plan to direct servo application– Fewer moving parts– More controllable

• Rotates 135 degrees and holds position• Goes down slowly, comes back quickly

Image Comparison

• Switched from C++ and Grayscale• Used Mat-Lab and Color– Built in image acquisition tools– Can interface easily with Arduino– No need to convert to grayscale• Performs calculations just as fast regardless

Image Comparison Cont.

• Get baseline image• Start video• Grab fames out of the video• Compare frames to baseline• Determine time for fingernail to return to

color

Validity of Tests

• 99% Reproducibility? Yes

Validity of Tests cont.

• Real value in 99% confidence interval• 99% probability that test is accurate• Unable to reject Null Hypothesis

• 99% Sure test is accurate

• Accurate within 3%? Yes

Questions?