{"id":"d992ee35-cdc7-46f0-a8c4-61ec6b975055","arxiv_id":"2607.08189","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"Closed-form input-constrained spatiotemporal tubes guarantee finite-time reach-avoid-stay for unknown Euler–Lagrange systems under actuator limits via offline feasibility conditions.","lead":"The paper gives a closed-form controller that keeps unknown robot dynamics inside a shrinking safe tube while never exceeding actuator limits, even with moving obstacles. It matters because most formal safety methods either need a model, online optimization, or ignore torque/force limits.","discovery_kind":"extension","skeptic_critique":{"model":"grok-4.5","headline":"No significant objection identified beyond the reader's already-correct identification of Assumption 5.","rationale":"The central claim of Theorem 4.3 is carefully conditioned on Ass. 1–5 and the offline inequalities (27)–(28). The only structural premise that can break the finite-time reach-and-stay half without violating the other hypotheses is Ass. 5, which the reader already flagged. The proofs of tube invariance (Thm. 3.3), radius positivity, and the two-stage contradiction arguments under the feasibility bounds are coherent; the controller is closed-form and input-saturating by construction of Ψ. Validation (simulations + hardware) is consistent with the theory when Ass. 5 holds. Because the paper itself treats Ass. 5 as an environmental premise rather than a controller-enforced property, the CONDITIONAL verdict already correctly reflects the scope of the guarantee. No stronger or independent load-bearing flaw was identified that would move the verdict.","tokens_in":21043,"tokens_out":536,"duration_ms":5381,"concrete_test":"Construct a 2-D scenario in which a single obstacle of velocity vo = k2 continuously orbits the target at distance < ra after t = 0 (so Ass. 5 never holds). Run the closed-loop controller of Theorem 4.3 with feasibility (27)–(28) satisfied; verify that x(t) stays inside Γ(t) and avoids U(t) for all t, yet never enters T. If the trajectory still reaches T in finite time, the paper's claim about the necessity of Ass. 5 is overstated; if it only avoids forever, Ass. 5 is confirmed as the precise boundary of the FT-RAS guarantee.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The reader's weakest_assumption (Assumption 5 / Remark 3.2) is already the correct load-bearing premise for the finite-time reach-and-stay half of Theorem 4.3. The paper is explicit that without eventual separation of unsafe sets from the tube center by ra, only perpetual avoidance is guaranteed; the goal-driven term never dominates and the finite-time claim of Definition 2.1 fails. The remainder of the argument (tube construction under Ass. 4, radius positivity, closed-form bounded controller under Ass. 1–3, offline feasibility (27)–(28), and Stage-1/Stage-2 contradiction proofs) is internally consistent and scoped correctly. No deeper hidden inconsistency or unstated assumption that would independently falsify the strongest claim was found.","agreement_with_reader":"agree"},"referee_report":{"model":"grok-4.5","summary":"The paper extends the spatiotemporal tube (STT) framework to unknown Euler–Lagrange systems under hard actuator limits, targeting finite-time reach-avoid-stay (FT-RAS) tasks in dynamic environments. It designs center and radius ODEs for a time-varying ball that is goal-directed and obstacle-avoiding, derives offline feasibility conditions relating available control authority, uncertainty bounds, and tube rates, and synthesizes a closed-form two-stage controller based on bounded transformations that keeps the state inside the tube while enforcing |τ|≤τ̄ by construction. Theorems 3.3 and 4.3 provide the main guarantees (tube FT-RAS properties under Assumptions 4–5; closed-loop tube invariance under feasibility (27)–(28)). Validation includes simulations on a mobile robot, quadrotor, and spacecraft, plus hardware experiments on a differential-drive robot, and a comparison against unconstrained real-time STT.","tokens_in":21316,"tokens_out":1194,"duration_ms":28945,"significance":"If the claims hold under the stated assumptions, the work addresses a genuine gap: formal FT-RAS for unknown EL systems with input constraints in dynamic environments, without models or online optimization. Strengths include approximation-free closed-form laws suitable for real time, offline-verifiable feasibility conditions that link actuator limits to tube design, structured contradiction/Lyapunov proofs, multi-platform simulation evidence, and hardware validation with explicit input-bound satisfaction. The quantitative comparison with unconstrained STT usefully illustrates the performance–authority trade-off. The contribution is incremental relative to the authors’ prior STT series but practically meaningful for safety-critical robotics under saturation.","major_comments":[{"comment":"Assumption 5 (and Remark 3.2) is load-bearing for the finite-time reach-and-stay half of Definition 2.1 / Problem 2.3 / Theorem 4.3: without a finite t1 after which all unsafe sets remain at least ra from the tube center, only perpetual avoidance is guaranteed and the goal-driven term never dominates. Theorem 4.3 currently states that the closed-form laws ensure FT-RAS whenever x(0)∈Γ(0) and (27)–(28) hold, without restating dependence on Assumption 5. The abstract and introduction similarly claim FT-RAS guarantees without flagging this environmental premise. The theorem statement (and abstract) should make the conditional nature of finite-time reach explicit, while retaining the unconditional safety claim.","section":"Assumption 5, Remark 3.2, Theorem 4.3, Abstract"},{"comment":"Lemma 4.1 and feasibility condition (28) depend on constants α, β, θ from Proposition A.1 (bounds on ∂Ψ/∂s, (∂Ψ/∂s)s, and Ψ(s)/s). The concrete transformation in Appendix A is only defined piecewise; the paper does not compute or bound α, β, θ for that map. Without explicit values or a short derivation, condition (28) is not fully constructive offline as claimed in Remark 4.2. Provide the bounds (or a procedure) for the chosen Ψ so that feasibility can be checked from the design parameters alone.","section":"Lemma 4.1, Eq. (28), Proposition A.1, Appendix A"}],"minor_comments":[{"comment":"Affiliations contain double commas (“Science„ Bengaluru”). Clean typography throughout.","section":"Title page / author block"},{"comment":"Notation drifts between x and x1 (e.g., vr(x1,t), e1(x1,t) in Stage 1) while the system state is x. Unify.","section":"Section 4.1"},{"comment":"Equation (4) for center dynamics is written with an informal case split; under Assumption 4 at most one j is active, but the typesetting would benefit from an explicit “let j* be the unique index with d(j*)≤ra” clause.","section":"Section 3.2, Eq. (4)"},{"comment":"Table 1 lists “Formal Guarantee” for the proposed method without noting that full FT-RAS is conditional on Assumption 5; a footnote would align the table with Remark 3.2.","section":"Table 1"},{"comment":"Figures 2–6 rely on external video links; ensure still frames and captions are self-contained for print (e.g., annotate obstacle velocities and tube radius evolution more clearly).","section":"Section 5"},{"comment":"A short design recipe for choosing k1,k2,k3, ra, rmin under (27)–(28) would help practitioners; currently parameters are stated case-by-case without a systematic selection procedure.","section":"Sections 3–4, Case Studies"}],"recommendation":"minor_revision","confidential_remarks":"The technical core is sound and the hardware experiment is a genuine plus. Novelty is real but incremental on the authors’ own STT line ([22]–[24]); the main delta is input-constraint compatibility plus feasibility conditions. Assumption 5 is the only structural limitation of the FT-RAS claim and is already disclosed, so minor revision (explicit theorem/abstract wording + constructive α,β,θ) should suffice. Scope fits eess.SY / control journals well."},"author_rebuttal":null,"desk_editor":{"model":"grok-4.5","letter":"This is a solid engineering extension of the authors’ own spatiotemporal-tube work, not a reinvention of the field. What is actually new is the input-bounded transformation in the two-stage controller plus the offline inequalities that relate available torque, uncertainty bounds, and tube design rates. Prior STT papers [22–24] did not enforce actuator limits; this one does, by construction, and keeps the controller closed-form and approximation-free.\n\nThe math is coherent and scoped correctly. Theorem 3.3 on tube construction and Theorem 4.3 on invariance under feasibility are standard Lyapunov/contradiction arguments done carefully. Lemma 3.5 and 4.1 give usable bounds. Multi-platform evidence (mobile robot sim + hardware, quadrotor, spacecraft) is better than most papers in this line, and the head-to-head with unconstrained STT makes the practical point: without the feasibility check you saturate and risk safety.\n\nSoft spots, in proportion. Assumption 5 is load-bearing for the reach-and-stay half: without an eventual free corridor to the target you only get perpetual avoidance. The paper is explicit about that; it is an environmental premise, not something the controller enforces. “Unknown” dynamics still need known bounds (m, mi, Vmax_M, d̄), so model-freeness is partial. There are many free gains. None of these falsify the claims under the stated assumptions; they mean this is a design-time tool, not a plug-and-play guarantee in arbitrary scenes. Self-citation is heavy but expected—the tube architecture is theirs.\n\nWho it is for: people building real-time safe controllers for EL robots who want formal FT-RAS under torque limits without online QP. Worth a serious referee. I would send it to peer review and would cite the feasibility conditions if I were writing on input-constrained prescribed-performance or STT-style tubes.","headline":"Clean, usable extension of the authors’ STT line that finally puts actuator limits and offline feasibility into a closed-form FT-RAS controller.","tokens_in":21927,"tokens_out":488,"would_cite":true,"duration_ms":16199,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"grok-4.5","headline":"Unknown robots with limited actuators can still meet finite-time reach-avoid-stay goals by tracking input-compatible spatiotemporal tubes.","keywords":["input constraints","safety guarantees","spatiotemporal tube","unknown Euler-Lagrange systems","finite-time reach-avoid-stay","approximation-free control","dynamic environments"],"falsifier":"Run the same mobile-robot hardware trial with actuator limits deliberately set below the offline feasibility bound; if the robot either saturates, leaves the tube, or collides while the mathematics claims the bound is violated, the central claim fails.","tokens_in":21896,"feed_emoji":"🤖","tokens_out":651,"duration_ms":6738,"temperature":0.7,"pith_summary":"When a robot’s equations of motion are unknown and its motors have hard force or torque limits, guaranteeing that it will reach a goal in finite time, avoid moving obstacles, and stay at the goal is hard. This paper shows that the problem can be solved by first building a moving ball (a spatiotemporal tube) whose path and shrinking radius keep the robot clear of obstacles and eventually inside the target, then applying a simple closed-form controller that never asks for more force than the actuators can supply. Offline inequalities link the available actuator strength, the size of unknown disturbances, and the speed of the tube so that, whenever those inequalities hold, the robot is mathematically forced to stay inside the tube. Because the controller is a handful of algebraic expressions with no online optimization or model learning, it runs in real time on ordinary robots. Simulations on a ground robot, a quadrotor and a spacecraft, plus hardware trials on a mobile robot, confirm that the vehicle stays safe and reaches its target while the commanded torques remain inside the prescribed limits.","feed_headline":"Limited actuators still meet finite-time safe navigation","feed_subtitle":"Closed-form tubes keep unknown robots inside actuator limits while reaching targets","key_machinery":"Input-constrained spatiotemporal tubes: a time-varying ball whose center and radius evolve according to explicit obstacle-avoidance and goal-seeking dynamics, together with a two-stage controller that uses bounded transformation functions so that both the virtual velocity and the actual torque always remain inside their prescribed limits.","core_discovery":"For an unknown Euler–Lagrange system subject to known actuator bounds, if the initial state lies inside a carefully designed spatiotemporal tube and two offline feasibility inequalities that relate actuator authority, uncertainty bounds and tube speeds are satisfied, then the closed-form velocity and torque laws keep the state inside the tube for all future time, thereby enforcing the finite-time reach-avoid-stay specification while never violating the input limits.","pith_inferences":[],"forward_implications":[],"fun_headline_variants":["Actuator bounds met via closed-form tubes for unknown systems","Offline conditions guarantee input-safe finite-time navigation","Spatiotemporal tubes keep limited actuators inside FT-RAS specs","Unknown EL systems stay safe under actuator limits with tubes","Real-time tubes enforce reach-avoid-stay without input violation"],"cache_read_input_tokens":16512,"weakest_assumption_plain":"The environment must eventually leave a clear corridor of fixed width around the tube center so that the goal-seeking term can take over; without that separation the method only guarantees perpetual avoidance, not finite-time arrival.","fun_headline_variants_meta":{"raw":{"variants":["Actuator bounds met via closed-form tubes for unknown systems","Offline conditions guarantee input-safe finite-time navigation","Spatiotemporal tubes keep limited actuators inside FT-RAS specs","Unknown EL systems stay safe under actuator limits with tubes","Real-time tubes enforce reach-avoid-stay without input violation"]},"model":"grok-4.5","effort":"low","cost_usd":0.006536,"raw_usage":{"total_tokens":1556,"prompt_tokens":700,"num_sources_used":0,"completion_tokens":85,"cost_in_usd_ticks":65360000,"prompt_tokens_details":{"text_tokens":700,"audio_tokens":0,"image_tokens":0,"cached_tokens":0},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":771,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":700,"tokens_out":85,"duration_ms":8746,"temperature":1.0,"reasoning_tokens":771,"cache_read_input_tokens":0,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-07-10T11:38:15.492423+00:00","model_set":{"reader":"grok-4.5"},"falsifier":"Run the same mobile-robot hardware trial with actuator limits deliberately set below the offline feasibility bound; if the robot either saturates, leaves the tube, or collides while the mathematics claims the bound is violated, the central claim fails.","supporting_citations":[],"review_version":1}