Reachability Analysis¶
Method — instantiates Phase-Space Mapping
Tests which states can be reached from current conditions under available controls and constraints.
Reachability Analysis is the one mechanism in this family that computes the set of states the system can be driven into — or kept out of — from where it is right now, given the controls actually available, the constraints in force, and a fixed time budget. It answers "can we get there from here?" and "can we still avoid that?", not "where does the system drift on its own." Its defining move is set propagation: it starts from the current state, applies every admissible control under the transition dynamics, and grows the frontier of what is reachable, pruning anything a constraint blocks. Run backward from a hazard, the same machinery yields the keep-out set — the states from which a bad outcome can no longer be avoided.
Example¶
An autonomous quadrotor is skirting the edge of a no-fly boundary with a depleting battery and a crosswind. The operator needs two answers before the next command. First, the forward reachable set thirty seconds out: given maximum thrust, turn rate, and remaining charge against the wind, does the set of states the drone can still occupy include a safe landing pad? Second, the backward reachable tube of the no-fly zone: is the drone already inside the set of states from which — under worst-case gusts — a boundary breach is unavoidable? The constraint boundaries here are hard: turn radius, climb rate, and battery. The forbidden region is the airspace it must not enter. Reachability returns not a trajectory but a verdict on possibility — landing is still reachable, breach is not yet unavoidable, and there are eleven seconds before that stops being true.
How it works¶
- Seed the frontier — begin at the current state (or, running backward, at the forbidden region's edge).
- Apply admissible controls — expand the frontier under the transition relation for every control the system can actually command, optionally against worst-case disturbance.
- Prune against constraints — intersect the growing set with capacity, resource, and legal limits; states past a constraint boundary drop out as unreachable.
- Mark the keep-out set — compute the states of no return around the forbidden region, so entry can be prevented before it is forced.
Tuning parameters¶
- Time horizon — how far ahead the set is propagated; longer horizons reveal more reach but accumulate approximation error.
- Control-and-disturbance model — optimistic (system chooses controls freely) versus adversarial (a disturbance fights back); the adversarial framing yields safe guarantees, the optimistic one yields optimistic reach.
- Over- vs under-approximation — a conservative outer bound never misses a reachable state but may cry wolf; an inner bound never overclaims but may miss a real path.
- State discretization — grid resolution or set representation; finer bounds cost compute and can dominate the run.
- Reversibility handling — whether transitions are treated as one-way, changing the shape of both the reachable and keep-out sets.
When it helps, and when it misleads¶
Its strength is killing false reachability — the target drawn on a map with no feasible path to it — and proving safety margins before they are tested. It is the mechanism that catches a point of no return while there is still time to steer, and that distinguishes "impossible" from merely "unlikely."[n1] It converts an argument about whether a state is attainable into a computation.
Its signature failure is approximation error pointed the wrong way: an over-approximation blocks feasible actions with phantom risk, while an under-approximation quietly certifies a route that a modeling gap makes unsafe. The classic misuse is treating "reachable in principle" as "reachable in practice" — the set says a state is attainable under the assumed controls, but says nothing about whether the organization has the resources, will, or coordination to command those controls. The guarding discipline is to state plainly whether a bound is inner or outer, and to validate the control-and-disturbance model against reality rather than assuming it.
How it implements the components¶
reachable_state_set— its primary output: the states attainable from current conditions under available controls within the horizon.constraint_boundary— the limits that prune the frontier; the analysis makes explicit which boundaries actually bind on reach.forbidden_state_region— running backward, it computes the keep-out set and the states of no return around a hazard.
It does not model where the system settles on its own (attractor_region, state_transition_relation) — that's Attractor Basin Analysis, its nearest twin: reachability asks where the system can be sent, basin analysis asks where it ends up unforced. It also does not vary assumptions to compare reachable sets across futures (trajectory_uncertainty_band, map_update_rule) — that's Scenario State Map.
Related¶
- Instantiates: Phase-Space Mapping — supplies the reachable-set and keep-out layer that separates feasible targets from fantasies.
- Consumes: Control-State Diagram — supplies the permitted transitions and controls the analysis propagates over.
- Sibling mechanisms: Attractor Basin Analysis · Behavioral State Space · Scenario State Map · Phase Space Plot · Trajectory Mapping Diagram · Risk Landscape Map · Control-State Diagram · State-Space Model
Editorial Notes¶
Form Classification¶
Form family: Analysis, Modeling & Optimization
Rationale: Reachability Analysis operates as an analytical, modeling, inference, comparison, or optimization procedure that derives insight or a solution because it tests which states can be reached from current conditions under available controls and constraints.
Independent corroboration: The frozen evidence defines Reachability Analysis as 'Tests which states can be reached from current conditions under available controls and constraints', so its operative form is Analysis, Modeling & Optimization.
Review outcome: Independent reviewer agreement; high confidence.
Origin Attribution¶
Primary origin: Engineering & Design
Origin pattern: Cross-disciplinary synthesis
Present-day reach: Multi-domain
Rationale: State reachability was canonically formalized in modern control engineering.
Related originating lineages:
- Mathematics — Linear algebra and dynamical-systems theory provide its formal tests.
- Systems Thinking & Cybernetics — Cybernetics materially shaped the control-and-state framing.
Review resolution: Both blind reviewers agree that engineering_design is the primary origin. Explicit reconciliation of alternate origin disagreement, domain reach disagreement adopts reviewer_a's classification because state reachability was canonically formalized in modern control engineering. The resulting lineage records alternates=mathematics, systems_cybernetics, origin_mode=cross_disciplinary_synthesis, and domain_reach=multi_domain; these describe formative provenance separately from later applicability.
Review outcome: Reconciled after independent review; high confidence.
Notes¶
[n1] The reachable set is a control-theory construct: the set of all states a system can attain from a given start under its admissible inputs and constraints, often computed via Hamilton-Jacobi methods or set propagation. Its backward form, the keep-out or unsafe set, underlies flight-envelope protection and collision-avoidance guarantees — cases where knowing a state is no longer avoidable is the whole point. ↩