Delivery Boi

• 

  Slack message about the delivery

Inspiration

Out office managers (Monika and Mikel spend a lot of time and effort to go around finding people and informing them about packages that have arrived for them. We felt we should automate this whole task and reduce the burden on Monika who is already doing so much for us.

What it does

Essentially a slackbot having a camera for eyes and Textract for a brain, that automates the whole process of detecting any newly arrived package, detects the name on its label, does a fuzzy matching with all the employee names in the office and sends them a personal message on slack to inform them about its arrival.

How I built it

Raspberry Pi 3B+ Object detection Optical character recognition (Textract and Tesseract) Slackbot

Challenges I ran into

Training the OCR model to recognize employee names (proper nouns) with various cultural backgrounds.

Accomplishments that I'm proud of

Automating the whole process and streamlining the employee search

What I learned

Various tools and packages for OCR, fuzzy matching (Benefits and disadvantages)

What's next for Delivery Boi

1. Voice-to-text capabilities to provide an additional feature to find and inform employees.
2. Better camera for the raspberry pi (we were recycling an old low resolution camera from our last hackathon)
3. A fancy casing and stand for the whole setup

Built With

• fuse.js
• fuzzymatching
• node.js
• opencv
• optical-character-recognition
• python
• raspberrypi
• slack
• tesseract
• textract
• vue