Mako Platform Reality

Stryker acquired Mako Surgical in 2013 and has since built the orthopedic robotics platform into a procedural anchor of its joint reconstruction business. Mako Partial Knee was the original indication, and the platform later added Mako Total Knee and Mako Total Hip, with Mako Total Knee widely reported as the volume driver. Stryker has publicly discussed extending the platform toward shoulder arthroplasty as part of its broader robotics roadmap. Across these indications Stryker has reported a substantial and growing installed base and cumulative procedure volume, and the competitive dynamic is reinforced by implant pull-through: a Mako system biases hospital purchasing toward Stryker's own implant families, such as the Triathlon knee. Any specific market-share, procedure-count, or clearance figures are public-domain context rather than claims of this filing; readers should confirm current figures against Stryker's own disclosures.

Mako is widely cited as holding a leading share of the US robotic joint replacement market. That leadership is the reason the platform is a useful reference point for a comparison about surgical actuation architecture, not a target for criticism. Mako does what it does well: it is a mature, cleared, surgeon-controlled system with a decade-plus safety and adoption record.

Operationally, Mako today is a haptically constrained cutting guide. The surgeon performs a CT-based pre-operative plan, registers the patient anatomy intra-operatively, and executes bone resection through a robotic arm that physically prevents excursion outside the planned envelope. The robot does not make autonomous cuts. It does not advance the burr or saw on its own initiative. It enforces a boundary, and the surgeon supplies the motive force. This is the regulatory and clinical equilibrium that the platform has occupied for a decade.

Path Toward Autonomous Phases

A broad direction for the category is to extend autonomy further into the procedure: assistance with registration verification, guided burr advancement within haptic envelopes, soft-tissue balancing assessment, and eventually more autonomous execution of well-circumscribed sub-phases. Each step toward autonomy raises a governance question distinct from the model-quality question. The FDA's Predetermined Change Control Plan framework is a real and relevant public mechanism for pre-clearing iterative updates to AI- and ML-enabled device software, but by design it addresses how a manufacturer may modify a model over time, not how an actuator's authority is delegated and revoked moment to moment inside a single case.

That gap is architectural, not a Stryker-specific fault. Nothing in a change-control plan, a surgeon training curriculum, or a post-market surveillance file is a machine-checkable predicate that a robotic actuator can evaluate at a sub-phase boundary to answer: under what conditions may this actuator advance toward more autonomous operation, and under what conditions must it retreat? Today that answer is encoded implicitly across documents and human judgment. No mainstream surgical robotics platform, Mako included, currently exposes it as a first-class, evaluated, auditable operand.

Governed actuation, as disclosed in the provisional, makes that encoding explicit. The disclosed graduated-actuation mode selector maps a composite-admissibility determination onto a bounded ladder of actuation modes rather than a binary permit-or-deny. In the spec these modes include disabled, simulated, advisory, consultative, shadowed, partial, constrained, stage-gated, deferred, full, and emergency-accelerated. Each mode is bound to admissibility predicates over credentialed observations, an authority taxonomy, freshness, policy, and the reversibility classification of the pending action. Bone resection is irreversible. Pin placement is partially reversible. Trial implant insertion is reversible. The admissible mode at any instant is the floor across these dimensions, and the disclosed evaluator recomputes that floor continuously rather than fixing it at submission time.

Architectural Fit for Mako Sub-Phases

A total-knee procedure decomposes into a sequence of sub-phases: exposure, array placement, registration, gap balancing, femoral resection, tibial resection, trial reduction, final implantation, and closure, with intermediate verification steps. Each sub-phase carries a distinct reversibility profile and a distinct admissibility envelope. The disclosed stage-gated mode lets an implementer bind each sub-phase to a gate whose entry condition is composite: the operator must hold the required authority credential, registration must hold to within a stated tolerance, the prior sub-phase must have closed cleanly, and the pending action must be classified at or below the reversibility ceiling for the currently admissible mode. The spec's stage-gated mode explicitly executes in a sequence of stages with admissibility re-evaluation between stages, enabling interruption or modification between stages.

Reversibility-aware commitment-point evaluation, which the spec discloses as an evaluator preferring reversible actuation paths where feasible, then governs what happens inside the gate. For an irreversible action such as the cut, the disclosed evaluator can hold authority with the surgeon and constrain the actuator to advisory or consultative operation regardless of planning-model confidence. For a reversible action such as a trial reduction, a more autonomous mode can be admitted under tighter telemetry because a wrong outcome is recoverable by retraction. The gradation is not a marketing label; it is a structural property of the action, and the disclosed architecture makes that property a first-class operand rather than a boundary drawn once at plan time.

The composition extends to multi-vendor operating rooms. When a robotic platform runs alongside navigation systems, intra-operative imaging, and implant components from several suppliers, an implementer needs a substrate that can admit cross-vendor observations without flattening the authority chain. Composite admissibility, in which each contributing system supplies credentialed observations and the evaluator weighs the composite against authority, freshness, and integrity, is the disclosed mechanism for integrating heterogeneous inputs without surrendering the actuation envelope. The spec's admissibility outcomes are not binary: they span admit, gate, defer, solicit, reject, and escalate, so a marginal or conflicting observation can trigger active corroboration or a mode retreat rather than a silent accept-or-halt.

Stryker Position

The category's competitive position rests on installed base and implant pull-through, but the medium-term architectural question is common to every vendor: as autonomy extends into the procedure, what expresses the authority model in a form a device and a regulator can both evaluate? Zimmer Biomet's ROSA, Smith and Nephew's CORI, and Johnson and Johnson's VELYS (DePuy Synthes) are each publicly advancing orthopedic robotics, and the navigation and imaging vendors around them are doing the same. The governed-actuation axis, composite admissibility, graduated modes, and reversibility-aware commitment, is orthogonal to which arm or which implant a vendor ships; it is a layer any of them, or an independent implementer, could adopt.

An implementer applying the disclosed approach would represent each candidate physical action as a proposed actuation carrying a reversibility classification, evaluate it through a composite-admissibility evaluator against credentialed observations and an authority taxonomy, select an actuation mode from the graduated ladder, and record the decision, the contributing factors, and the governance-policy version in a lineage field. This converts an implicit authority model into an explicit, auditable, machine-checkable artifact, which is the form regulators are increasingly likely to want for autonomous-phase operation.

Embodiments and Variations

The disclosed approach is not limited to orthopedic surgery. A skilled implementer could apply it to any actuator whose actions carry differing reversibility: a burr, saw, or drill in surgical robotics; a steering, braking, or throttle command in a self-driving or driver-assist vehicle; a flight-control surface or thrust command in an eVTOL; a bucket, blade, or boom in mining, construction, or agricultural autonomy; a gripper or arm in an industrial cobot; or an infrastructure actuator such as a signal phase or valve state. Variations disclosed or contemplated include: mode ladders configured per actuator class; admissibility predicates combining any subset of authority credential, freshness, corroboration, integrity, and reversibility; deferral with an expiration parameter that resolves to admit, gate, or reject; solicitation that emits a governed discovery query to gather corroborating observations; emergency preemption bounded by a preemption budget and expiration; graceful degradation that steps the mode down under reduced confidence rather than halting; and lineage recording that spans single-agent, distributed, and hybrid topologies. The mechanism is independent of any specific signaling medium, authority taxonomy, or deployment domain, and an implementer of ordinary skill in control systems and medical or autonomous-device software could build it from the disclosure.

Disclosure Scope

The invention described here, governed actuation as a composite-admissibility-evaluated, reversibility-aware, graduated, lineage-recorded, revocable actuation layer, is disclosed in U.S. Provisional Application No. 64/049,409. All statements about the platform's mechanisms, modes, admissibility outcomes, and guarantees trace to that filing. All references to Stryker, Mako, Zimmer Biomet, ROSA, Smith and Nephew, CORI, Johnson and Johnson, DePuy Synthes, VELYS, the FDA, the Predetermined Change Control Plan framework, market-share figures, procedure counts, implant families, and clearance status are external market and regulatory context, provided for positioning only. They are drawn from public reporting, are not claims of U.S. Provisional Application No. 64/049,409, and should be verified against the respective companies' and agencies' own current disclosures. No affiliation, endorsement, incident, contract, or capability beyond what those parties have publicly stated is asserted or implied.