P19095: LiveAbility Lab
/public/

Detailed Design

Table of Contents

Team Vision for Detailed Design Phase

The focus of this phase was getting caught up. Following our preliminary detailed design review we realized that our progress was behind. The primary focus of this phase was completing items that should have been completed prior to the previous phase as well as advancing the project to having the ability to confidently order all components so that assembly will begin at the start of the semester. Areas of notable progress include: accelerometer testing, A301 force sensor circuit built, system architecture simulated, and a proof of concept alternative to the A301 sensors. At this time, all major components for the system to analyze force and location for a user who does not utilize an assistive device have been ordered. Accelerometers and minor electrical components still need to be ordered.

Prototyping, Engineering Analysis, Simulation

Iterative activities to demonstrate feasibility, including assumptions you made in your analyses or simulations. Have you completed sufficient analysis to ensure that your design will satisfy requirements? Have you included all usage scenarios in your modeling?

Accelerometer Evaluation

Thanks to Dr. Puchades, the team was able to test out some functionalities of the MMR accelerometer before purchasing. Location calculation and typical output were two areas of interest.

A study was done into the feasibility of calculating location using an accelerometer. Previous research performed by Dr. Puchades indicated that over 100m of error was introduced in a short time frame. The setup of the experiment was to have two accelerometers measure for a short time while laying still. Once the data was exported, two methods of location calculation were used; Excel and MATLAB. The idea behind using two accelerometers is that if the location error given in both cases was constant, then this is representative of noise which can be filtered out of the system. The analysis of the data using Excel and a rectangular approximation found that the error in the location was not a constant noise value, each accelerometer reportedly traveled hundreds of meters in the course of a few minutes.

The software that is included with the MMR accelerometer can provide a data output which can be plugged into Matlab. The software is very user friendly and does not need to be recreated for the usage in our device. For each patient, a normal walking pattern will need to be established so that machine learning can be utilized to create bins. These bins will be included in the profile of a patient. When the patient uses the accelerometers for extended use, the data will be filtered by the bins, abnormalities can be analyzed individually to diagnose and treat specific problems.

Accelerometer Still Raw Data

Accelerometer Still Raw Data

Accelerometer Moving Raw Data

Accelerometer Moving Raw Data

Accelerometer Still Test

Accelerometer Still Test

Accelerometer Still Test Zoomed In

Accelerometer Still Test Zoomed In

Accelerometer Moving Test

Accelerometer Moving Test

Accelerometer Moving Test Zoomed In

Accelerometer Moving Test Zoomed In

Fabrication of a Capacitive Force Sensor

An idea that came out of the last phase was to fabricate our own force sensors in order to simplify the force measurement process, add durability, and save money. A simple force sensor can be created by making a parallel plate capacitor with a rubbery dielectric between them. The capacitance will increase as the space between the plates decreases. In this phase a simple parallel plate capacitor was created to assess the feasibility of creating a similar device.
Parallel Plate Capacitor Theory

Parallel Plate Capacitor Theory

Brazing the Capacitor

Brazing the Capacitor

Completed Silicone Capacitor

Completed Silicone Capacitor

Capacitor Demo: https://youtu.be/pJu7p9NzeGE

The results are encouraging. Moving forward, Nick will take ownership of the homemade force sensors, working into the next semester to create a product to rival the A301.

PCB Layout for our A301 Op-Amp Circuitry

A301 Recommended Op-Amp

A301 Recommended Op-Amp

Schematic of the PCB on Circuitmaker

Schematic of the PCB on Circuitmaker

This is the basic op-amp circuit needed to work with the A301 pressure sensors to gather the data. Along with the op-amp circuitry, there is a voltage divider to supply a reference voltage to the A301 sensor itself and also an A/D Converter. This A/D Converter is not tied to anything yet because pins still have to be chosen and finalized.
2d representation of the PCB Layout

2d representation of the PCB Layout

This is a model of the 2d PCB layout. Once, the A/D Converter circuitry is finalized to communicate with the Pi, the PCB can be further manipulated to save space and cost.
3d representation of the PCB Layout

3d representation of the PCB Layout

This is a current 3d model of the PCB Layout board.

Drawings, Schematics, Flow Charts, Simulations

Full System Schematic, Resistive Sensors

Full System Schematic, Resistive Sensors

This is a full system schematic to represent how our system will be connected to each other. The circuitry is not what is actually represented by our components since we can't model the full MMR Accelerometer or JuiceBox Zero Schematic. However, components and circuits are there to represent what each component and system will be. For example, the Juicebox zero is modeled by a buck converter because that is one of the parts of its schematic, but it also contains regulators and charging circuitry as well.

While initial testing of the A301 Resistive Pressure sensors indicated that a reference voltage of 3.7 would be the most effective, this schematic shows the reference at less than 3.3V. This was done to illustrate the Zero Pi's capacity to supply a reduced voltage to the pressure sensors, as may need be done with A301 sensors with more appropriate calibration.

Subsystem Wiring Instructions

Systems Architecture

Systems Architecture

Wiring for the MMR

Wiring for the MMR

Wiring for the decwave development board

Wiring for the decwave development board

Wiring for the juicebox

Wiring for the juicebox

Wiring for the battery subsystem

Wiring for the battery subsystem

Wiring for the raspberry pi

Wiring for the raspberry pi

Wiring for the A301

Wiring for the A301

Wiring for the created PCB

Wiring for the created PCB

Pinout for the Gateway

Pinout for the Gateway

Component Housing and Shoe Attachment

The below images show our concept for the housing for all components that will need to be attached to the individuals shoe. The only exception is the accelerometer which we are creating as an accessory to this main system, since it will only be used when gait analysis is desired. This housing will be made out of PLA plastic and 3D printed at the RIT construct. It will have a weight of 60.85 g (0.13 lbs) which will cost $1.83 to print. The total weight of the housing and all components will be about 0.45 lbs, so it will be lightweight and hopefully will have a negligible affect on the individuals natural gait.

Component Housing Top View

Component Housing Top View

Component Housing Iso View

Component Housing Iso View

Assistive Device Force Sensor Integration Concept

In terms of the assistive device attachment, our current solution is to create our own version of the rubber tips that are commonly found on the legs of assistive devices such a walkers, crutches, and canes, and create a slit for the force sensor to slide into. This will then be wired to a similar housing as shown above, which will be connected to the leg of the device. This is a preliminary design, our current focus is fully implementing the system to an individuals foot, but many of the same concepts can be applied to the device attachment.

Assistive Device Force Sensor Attachment

Assistive Device Force Sensor Attachment

Assistive Device Force Sensor Attachment Top View

Assistive Device Force Sensor Attachment Top View

Interfacing Sensors with Microcontroller and Networking

Interfacing sensor with microcontrollers and Networking

Bill of Material (BOM)

BOM

BOM

Test Plans

Test Plan list

Test Plan list

Battery Integrity Test - Simple Discharge Circuit

Battery Integrity Test - Simple Discharge Circuit

To test the integrity of the battery system, the battery will be charged utilizing the raspberry pi juicebox, then rapidly discharged near to the battery's maximum rate with the above circuit. This will be done ten times in order to gauge if the batteries can sustain worst case scenario abuse without compromise.

The total advertised capacity of the system is 6800mAh, but the system is only expected to need 450mA per hour (2.5mA per pressure sesnor; 100mA for accelerometer; 250mA estimated for Zero W; 100mA for RF Beacon). For eight hours, the battery needs to supply 3600mAh, and these batteries were chosen due to their optimal shape and capacity.

Risk Assessment

Risk Assessment Part 1

Risk Assessment Part 1

Risk Assessment Part 2

Risk Assessment Part 2

Plans for next phase

Team Goal for next semester

The team goal for the next phase is to assemble the components into a robust system. By the start of the next semester all components will be in our possession and we will piece them together. An important goal for next semester will be to establish a good line of communication with our industry experts who can give advice regarding design choices as well as offer test subjects.
Gantt Chart

Gantt Chart

Individual Responsibilities to Achieve Team Goals

Martine Bosch
Nick Petreikis
Matthew Devic
Patrick Mylott
John LeBrun
Hrishikesh Moholkar
Erik Brown

Home | Planning & Execution | Imagine RIT

Problem Definition | Systems Design | Preliminary Detailed Design | Detailed Design

Build & Test Prep | Subsystem Build & Test | Integrated System Build & Test | Customer Handoff & Final Project Documentation