Under the Bayh-Dole Act (35 U.S.C. §§200-212), a contractor that patents an invention made under federal funding — the routine case for NASA-funded ECLSS (life-support) hardware — keeps title to the patent, but the government retains a permanent non-exclusive license, and march-in rights: if the contractor has not taken effective steps to achieve "practical application" within a reasonable time, the agency can force licensing to third parties.
dP/dt = k(E) · P · (1 − P) k(E) = 0.15 + 0.9·E (logistic commercialization)
dM/dt = +18·(t − R)·(Pth − P) while t > R and P < Pth (risk accrues)
dM/dt = −0.6·M otherwise (risk decays once safe)
X(t) = max(0, 1 − t/20) (20-yr statutory term erosion)
X ← min(X, 0.15) permanently once M ≥ 100 (march-in invoked)
- Commercialization effort E — investment/marketing intensity the contractor puts in; raises the logistic growth rate k of practical-application progress P.
- Reasonable-time window R — how many years the funding agency allows before lagging progress starts accruing march-in risk.
- Practical-application bar — the progress level Pth the agency treats as "available to the public on reasonable terms."
- If the risk score M reaches 100, march-in is invoked: exclusivity collapses to the statutory license floor and outside licensees (the orbiting red markers) gain access — once invoked it is permanent, even if P later climbs past the threshold, mirroring the real, rarely-used but real, government clawback.
Numerically verified: the accrual/decay pair and the term-erosion formula reproduce the source model exactly — at low effort (≤0.3) march-in is invoked around year 7–9 even though P eventually reaches ~100%, confirming march-in is a one-way, pace-dependent clawback rather than a simple threshold check.
This 2D view swaps the 3D torus-knot module for a projected orbital dashboard: drag the main panel to rotate the camera around the patent core, exactly as the orbit controls did in 3D, and watch the strip chart below trace P, M and X live against the timeline.