Skip to content

TU-Munich/TechChallenge-Apoplex

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

1 Commit
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

TECH CHALLENGE - Apoplex

http://techchallenge.de/

INTRODUCTION

Three hours. That is the time you have to get help, if you just suffered a stroke. However, only 20% of the people reach the hospital in time. The consequences of being late can be remaining brain damages or even death. We couldn't accept this low rate of 20% and thought of a potential solution.

Apoplex is an Alexa skill that helps with the recognition of early-warning symptoms of strokes. Early recognition of stroke symptoms is crucial in order to seek help and undergo therapy as fast as possible.

The Apoplex skill was programmed in NodeJS, runs on the Amazon Echo Show device and uses the AWS Lambda and DynamoDB services.

THINGS USED IN THIS PROJECT

  • Amazon Alexa
  • Echo Show
  • Alexa Skill Builder
  • AWS Lambda
  • AWS Dynamo DB
  • AWS S3
  • AWS CLI
  • NodeJS
  • Sample NodeJS Quiz Game

(Github Link: https://github.com/alexa/skill-sample-nodejs-quiz-game)

THINGS NEEDED FOP THE PROJECT

  • Echo Show
  • AWS Account
  • Amazon Developer Account

Echo Show

Buy it via Amazon http://amzn.to/2DOh0G6

Amazon AWS Account

Register on https://aws.amazon.com

Amazon Developer Account

Register on https://developer.amazon.com/de/

RESOURCES

Questionnaire

The user has to answer the predefined questions in the questions/questions-questionnaire.json with either YES or NO.

Example dialogue:

Alexa: "Do you have a headache?"

Echo Show shows the question an the possible answers

User: Touches on YES button on the screen or says "YES"

Select Photos

The user has to select the right image out of 4 images via the Echo Show touch display. The question sequence can be found in the questions/questions-select.json file. The photos can be found in the `assets folder.

Example dialogue:

Alexa: "What photo represents a cat?"

Echo Show shows 4 photos. Dog, Cat, Mouse, Bird

User: Touches on cat photo on the screen

Describe Photos

The user has to describe in a word what he or she sees on the display. The question sequence can be found in the questions/questions-select.json file. The photos can be found in the assets

Example dialogue:

Alexa: "What object can you see in the photo?"

Echo Show shows a photo of a hammer

User: "Hammer"

VRIABLES

There are the following variables to control the skill.

Threshold of the max wrong answers
If there are more wrong answers than the threshold the call a doctor
EMERGENCY_QUESTION_THRESHOLD
EMERGENCY_SELECT_THRESHOLD
EMERGENCY_DESCRIBE_THRESHOLD
EMERGENCY_UNDERSTANDING_THRESHOLD
How many questions of the questionnaire should be asked
MAX_QUESTIONS 
How many image select questions should be asked
MAX_SELECT_QUESTIONS
How many image describe questions should be asked
MAX_DESCRIBE_QUESTIONS
This is the welcome message for when a user starts the skill without a specific in
WELCOME_MESSAGE
This is the message a user will hear when they start a quiz.
START_QUIZ_MESSAGE 
This is the message a user will hear when they are in the select section of the sk
START_QUIZ_SELECT_MESSAGE 
This is the message a user will hear when they are in the describe section of the 
START_QUIZ_DESCRIBE_MESSAGE
The user will hear the emergency understanding message if they have said not yes o
EMERGENCY_UNDERSTANDING_MESSAGE
This is the message a user will hear when they try to cancel or stop the skill, or
EXIT_SKILL_MESSAGE 
This is the message a user will hear after they ask (and hear) about a specific da
REPROMPT_SPEECH
This is the message a user will hear when they ask Alexa for help in your skill.
HELP_MESSAGE 
Background image fallback
IMAGE_FALLBACK

#Deploy to AWS Lambda Works with AWS CLI

sh publish.sh

#Images The following images have been used in the application.

https://stocksnap.io/search/microphone
https://stocksnap.io/photo/R8SYL6ROA1

Todo

  • Multi Lang
  • Emergency Interaction (call emergency, call relative, ...)

Team

Martin Schleich   

Jasmin-Chiara Bauer

Yasmin El Khodary

Sebastian Erhardt