Continuity Is the Capability: A Fault-Aware, User-Controlled Input Bridge for Accessible Computer Interaction
Year: 2026
Type: article
Abstract: Accessible computer input is not simply a matter of mapping a head motion or pneumatic pressure change to a cursor command. This technical report presents a small, dependency-free software state-machine prototype that tests input continuity: it takes timestamped head-motion, joystick, pneumatic, current, and explicit human control events; proposes dry-run HID events; and enforces fail-closed transitions for stale evidence, overcurrent, STOP, and ambiguous pneumatic measurements. On 1,000 deterministic, generated fault cases it preserved the predeclared no-unsafe-proposed-click invariant in 1,000/1,000 cases (709 head control, 150 joystick fallback, 141 hold/deny). Twenty-seven portfolio unit tests pass in normal and optimized Python. Software prototype only. No hardware was assembled, no USB HID was emitted, no users were studied, and no clinical or efficacy claims are made. The 1,000 constructed cases are implementation regression evidence, not real-world reliability statistics. Head tracking and sip-and-puff input are established prior art (including Makers Making Change LipSync); this report claims no invention of them. Reproduce: python3 -m unittest discover -s tests -v; python3 -O -m unittest discover -s tests -v; python3 run_portfolio.py; python3 demo.py. Python 3.10+ standard library only, no network or hardware needed. Files: the paper (PDF) and OPENLINE_INVENTION_PROTOTYPES_20261117.zip (392,793 bytes, SHA-256 09e3e9f7dab12b12561bf4309aafbb3ef91f520e2e7dcf522a1a91b776c2232e), the full reproducible bundle with source, tests, figures, and receipts. Public source: https://github.com/terryncew/openline-invention-prototypes Licenses: manuscript CC BY 4.0; original executable code Apache License 2.0.
Author: White, Terrynce Anson
Language: English
Subfield: Human-Computer Interaction
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7235192812
