Folk.computer

• Tabs
• Files
• Notifications
• Tweets
• Songs

• Fiducial Marker - QR codes
  • disadvantages:
    • pretty large
    • they have to be on camera! (occlusion)
  • ML Object recognition
  • OCR
  • M5 stick pi - $15, but need to charge
  • AirTags -
  • NFC tags - you can't track them far away, but cheap!
  • UHF RFID - readers aren't cheap, 96-bit
    • power over the radio rig

• easy to make new file, new tab
  • keep them small

• cheap
• small
• unique ID
• battery-free
• location sense - relative to others
• ideally has a "button" - activated or not

you want to physicalize your computational space

you're in a physical space and I want to colocate computations within that space

https://clementzheng.info/Printed-Paper-Markers

Embodied computer

RFID project (UHF)

• RFID
• can tell if they are inside an area or not
• range is long: 10-20 feet
• but sensing is not precise enough:
  • "is this thing in a bucket"
  • "did you pass through this gate in a warehouse"
• State of the art for RFID
  • track a tag with latency of 7ms
  • xyz to 1cm accuracy
  • 10 ft
  • "can we get this state of the art performance with $500-600 of gear (open source hardware)??"
    • still expensive, but we're making a computer
  • There are regulatory limitations around

what are the capabilities of a spatial computer that would you like to have?

• screenshots printed out quickly

general goal: spatial operating system

• have built-in primitives for running programs present in physical space that can be interacted with
• motivation for RFID: spatialize objects into the environment
  • Bookshelf in the system that the computer knows about.
  • Provide a new technological primitive

capabilities that happen to be available in the system limit what you with it

track interactions and

broader aspirations - the chess moves after that