{"id":"d578b6b4-e083-4803-9349-23c819531845","arxiv_id":"1908.02090","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"NAVARO II is a new 3-RPR parallel robot whose scissor mechanisms act as prismatic joints, and its eight actuation modes have compact singularity equations that allow singularity avoidance across the workspace.","lead":"This paper describes a new robot arm design that uses scissor mechanisms as sliding joints, giving the arm eight ways to switch which joints are powered. The authors derive equations for the arm's singular positions and show they can be avoided by switching power modes, which could allow a larger usable workspace.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The scissor-to-prismatic equivalence is not derived; the mapping from screw position μ_i to leg length ρ_i could introduce transmission singularities or multiple solutions that invalidate the computed singularity surfaces and the unique-IK claim.","rationale":"After reading the manuscript in good faith, the most load-bearing condition for the central claim is the assumed exact equivalence between the scissor mechanism and an ideal prismatic joint of variable length ρ_i. This is the premise on which the 3-RPR kinematic model, the one-solution IK statement, the Jacobian (Eqs. 5–6), the workspace limits (Eqs. 13–15), and the eight singularity surfaces (Eqs. 31–38) all rest. The paper never provides the internal kinematics of the scissor leg: it only states that ρ_i is computed from the screw position μ_i and bar length l. That omission is significant because the actual actuated coordinate is μ_i, not ρ_i. If the mapping is nonlinear or non-injective, the physical robot has additional singularities (where f'(μ_i)=0) and possibly multiple actuator configurations for the same pose, so the reported singularity surfaces and the unique-IK claim would not describe the built mechanism. The same concern was highlighted by the reader as the weakest assumption, and I agree. The path connectivity claim is secondary; even if the singularity surfaces were correct, non-superposition alone does not prove workspace traversability, but that is a separate issue. No change to the reader's conditional verdict is required: the paper should be accepted only if the scissor kinematics are supplied and the transmission mapping is shown to be singularity-free and injective over the design range. A single loop-closure calculation of one scissor leg would settle the concern.","tokens_in":10290,"tokens_out":11347,"duration_ms":107602,"concrete_test":"Derive the loop-closure equation of one scissor leg in Fig. 3: express the distance ρ_i = |A_iB_i| as a function of the screw position μ_i and the scissor bar length l. Then (a) verify that B_i moves along the straight line A_iB_i as μ_i varies, with no parasitic out-of-plane or rotational DOF; (b) compute f'(μ_i) over the range [rmin,rmax] used in Sec. 3.4 (rmin=8, rmax=59) and check whether f' vanishes at any interior point; (c) check whether f is injective on that interval. If f' vanishes inside the workspace or f is not injective, recompute the singularity surfaces with the actual actuator Jacobian and compare with Appendix Eqs. (31)–(38).","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central claim is that each scissor leg can be treated as an ideal prismatic joint of length ρ_i in the 3-RPR loop equations. Sections 3.1–3.4 and the Appendix derive Jacobians, workspace boundaries, and the eight singularity equations from that premise. But the physical actuator is the screw position μ_i; ρ_i is only said to be 'computed' from μ_i and the bar length l (Sec. 2). If ρ_i = f(μ_i), then the true actuated rates are μ_dot and the Jacobian factorization is A t = B diag(f'(μ_i)) μ_dot. The singularity set then includes the transmission surfaces f'(μ_i)=0, which are absent from Eqs. (31)–(38). Scissor mechanisms typically have zero transmission ratio at their closed and open limits; if such a configuration lies inside the rmin≤ρ_i≤rmax range used for the workspace, the physical singularity loci differ from those computed. In addition, the unique-IK statement holds for the coordinate ρ_i; in terms of the physical μ_i it requires f to be injective, which is never stated or shown. Because all later results depend on the exact equivalence ρ_i ↔ a translation along the leg axis, an unmodelled transmission singularity or non-injective mapping would break the main contribution.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper introduces NAVARO II, a planar 3-RPR parallel robot whose three prismatic joints are replaced by scissor mechanisms mounted in planes orthogonal to the motion plane. Each leg can be actuated either at its base revolute joint (angle theta_i) or at the scissor-driven prismatic coordinate (length rho_i), giving eight actuation modes selected by dual clutches. The authors write the loop-constraint equations for a symmetric equilateral base and platform, define the workspace boundary by rmin <= rho_i <= rmax, derive scissor dimensions for a prescribed regular workspace, and present compact algebraic singularity surfaces for all eight actuation modes. They claim that the 3-RPR kinematics gives exactly one inverse-kinematic solution in every actuation mode, and that because the eight singularity surfaces are not superposed, the whole workspace can be traversed without singularities by switching actuation modes.","tokens_in":10595,"tokens_out":8587,"duration_ms":87305,"significance":"If the central modeling equivalence between the scissor mechanism and an ideal prismatic joint is valid, this is an attractive mechanical realization of variable actuation: it retains the unique inverse-kinematic solution of the 3-RPR, replaces bulky prismatic actuators with compact scissors, and provides a single algebraic description of the parallel-singularity surfaces for all actuation modes. The explicit singularity equations in the Appendix are a potentially valuable reference, and the numerical design procedure for the scissor dimensions is straightforward to reuse. The paper does not rely on fitted parameters or circular reasoning; the constraint equations (7)-(12) are consistent with the stated geometry. However, the two load-bearing claims -- the scissor-to-prismatic equivalence and the path-feasibility inference from non-superposed singularity plots -- are not established by the current manuscript.","major_comments":[{"comment":"The kinematic model treats rho_i as an independent, ideal prismatic-joint coordinate: the Jacobian factorization (5) uses q_i = rho_i, and the workspace boundary is rmin <= rho_i <= rmax. However, the physical actuator is the screw position mu_i, and the manuscript only states that 'the length of the scissor rho_i is computed thanks to the position of prismatic link shaft mu_i and the length l of the scissors' without giving the mapping rho_i = f(mu_i). If f has transmission singularities f'(mu_i) = 0 inside the allowed range, or if f is not injective on that range, then the true singularity surfaces include extra factors and the claimed unique inverse-kinematic solution holds only in the rho-coordinates, not in the actuator coordinates. Because every later result depends on this equivalence, the authors should derive f, state its domain, prove injectivity and f' != 0 over [rmin, rmax], or incorporate the transmission Jacobian into the singularity analysis.","section":"Section 2 / Section 3.1, Eq. (5)"},{"comment":"The sentence 'As none of them are superposed, it is possible to completely move through the workspace by choosing a non singular actuation mode for any pose of the mobile platform' is not a valid inference. Non-superposition of the eight singularity surfaces only says that no two surfaces coincide everywhere; it does not imply that the union of the surfaces is not a separating barrier, nor that a continuous path avoiding all of them exists between arbitrary configurations with feasible actuation-mode switches. The later remark that an 'unrepresented singularity exists for alpha = pi' in mode 8 further complicates the picture. The authors should either provide a topological proof of path-connectedness of the singularity-free union of the mode regions, or demonstrate the mode-switching path planner on representative trajectories inside the workspace.","section":"Section 4, paragraph after Figs. 10-11"},{"comment":"The eight singularity equations are given without derivation, and no elimination script, intermediate polynomial system, or verification procedure is provided. Since these equations are the main technical result supporting the non-superposition and traversability claims, the reader cannot independently confirm that the displayed compact forms are correct or complete. The authors should supply the Maple/Siropa computation or an independent verification method, and should state explicitly for which portions of the workspace the algebraic elimination is valid, including any exceptional configurations such as alpha = pi in Eq. (38).","section":"Appendix, Eqs. (31)-(38)"}],"minor_comments":[{"comment":"The caption reads 'The eight actuating modes of the 3-RRR VAM', but the robot and table describe a 3-RPR variable-actuation mechanism; this appears to be a copy-paste error from the NaVARo I paper.","section":"Section 2, Table 1 caption"},{"comment":"The definitions of h_i for eliminating theta_i vs rho_i are mathematically consistent, but the sentence 'defined, for dot theta_i as h_i = (b_i - a_i)' is easy to misread as applying when theta_i is the actuated joint rather than the passive joint that must be eliminated; rewording would improve clarity.","section":"Section 3.1, Eqs. (3)-(4)"},{"comment":"The relation h = 9/n and l = sqrt(6481)/n is presented without comment on the units or the approximation steps; the derivation from rmin = 9.64 and rmax = 79.77 to the rounded values 9 and 80.5 should be stated explicitly.","section":"Section 3.4, Eq. (30)"},{"comment":"The sentence 'Figures 12 represents the singularities of the first actuation mode' has subject-verb disagreement, and the figure caption does not identify which panel corresponds to the workspace-limited case; adding labels would help.","section":"Section 4, Fig. 12"}],"recommendation":"major_revision","confidential_remarks":null},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper is a genuine design contribution: it puts scissor mechanisms on a 3-RPR variable-actuation topology, giving eight actuation modes and explicit singularity surfaces for all of them. The loop-closure equations, Jacobian, workspace boundary formulas, and scissor sizing equations are all written out cleanly, and the authors are honest about what is computed and what is only conjectured (e.g., stiffness). That alone makes it worth a careful read for anyone working on parallel mechanisms with variable actuation.\n\nThe reader's take and the stress-test note land on the real soft spot: the paper never defines the mapping from screw position μ_i to effective leg length ρ_i. It says only that ρ_i is \"computed\" from μ_i and the bar length. If that mapping has a vanishing derivative inside the operating range, the serial singularity set changes: you'd have f'(μ_i)=0 surfaces that are not in the appendix equations. The parallel singularities (det A=0) are geometric and probably survive, because they depend on the constraint structure, not the transmission. But for control and for the claim of one-solution IK, you need to know that f is monotone and f' ≠ 0 in [rmin, rmax]. That is a missing piece, not a demolition; a standard scissor has zero transmission ratio only at its extreme closure, which the authors already exclude by choosing rmin > 0. Still, they should state this and give the formula.\n\nThe second overstatement is the inference from non-superposed singularity plots to full workspace traversability. Non-overlap does not imply the union of the surfaces fails to cover the workspace, and even if every pose has at least one non-singular mode, you need a path that lets you switch modes without hitting a singularity. They cite a previous algorithm, so this is probably an editorial shortening, but as written the claim is too strong.\n\nThe paper's positive side: the singularity equations for all eight modes are a concrete, checkable result, and the workspace construction with a prescribed regular workspace is practical. No code, no prototype, no measurements, but the kinematic core is coherent.\n\nFor a reader: it's for robotics kinematics specialists. I'd send it to a serious referee; the requested revisions are to add the transmission model and to temper the traversability claim. With those, it's a solid journal paper.","headline":"A credible kinematic design paper for a scissor-based 3-RPR; the main caveat is the missing transmission model, not the core geometry.","tokens_in":11145,"tokens_out":6380,"would_cite":true,"duration_ms":71076,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["70B15"],"pacs":[],"model":"deepseek-v4-flash","headline":"Scissor legs give a planar parallel robot one inverse-kinematics solution per actuation mode and closed-form singularity equations.","keywords":["variable actuation","parallel robot","3-RPR mechanism","scissor mechanism","singularity surfaces","inverse kinematics","workspace design","actuation modes"],"falsifier":"Take the eight determinant equations from the Appendix and search for a real pose $(x, y, \\alpha)$ inside the nominal workspace that satisfies two different mode equations simultaneously; any such pose is a configuration where the claimed strategy cannot find a non-singular actuation mode. The same check could be run on a physical prototype by commanding a path through such a pose and observing whether the platform stiffens or drifts.","tokens_in":10064,"feed_emoji":"⚙️","tokens_out":8318,"duration_ms":79233,"temperature":0.7,"pith_summary":"The paper introduces NaVARo II, a planar parallel robot whose three legs are built with scissor mechanisms mounted vertically, so each leg acts as a variable-length sliding joint in the plane of motion. The robot has eight actuation modes, depending on whether each leg is driven at its base revolute joint or at its scissor extension. The paper's central claim is that, unlike the earlier 3-RRR (three-revolute-joint leg) version, this 3-RPR-based design has exactly one solution to the inverse kinematic problem in every actuation mode, and that the parallel-singularity surfaces for all eight modes can be written as compact algebraic equations. Inside the nominal workspace these surfaces do not coincide, so a controller can switch actuation modes and avoid singular configurations throughout the workspace. This matters because the robot needs only three motors plus six clutches instead of added actuators, and the vertical scissors give out-of-plane stiffness and a larger workspace than the parallelogram-based predecessor.","feed_headline":"Eight actuation modes let a planar robot dodge every singularity","feed_subtitle":"Scissor legs behave like sliding joints, so every actuation mode has one inverse solution and compact singularities.","key_machinery":"The mechanism that carries the argument is the scissor-as-prismatic-joint leg: a scissor lift mounted vertically, with extension $\\rho_i$ computed from the position of the prismatic link shaft and the scissor bar length, which is kinematically equivalent to a sliding joint in the planar 3-RPR constraint loop. The algebraic machinery is the loop-closure and pose equations (Eqs. 7-12), the direct and inverse Jacobian matrices $A$ and $B$, and an algebraic elimination step that turns $\\det(A)=0$ into the compact polynomial singularity equations for the eight modes in the Appendix. The workspace boundary is then defined simply by $r_{\\min} \\le \\rho_i \\le r_{\\max}$, which lets the authors choose scissor dimensions (bar length $l$, height $h$, number of scissors $n$) to enclose a prescribed regular workspace.","core_discovery":"The core discovery is that a scissor mechanism placed in the plane orthogonal to the robot's motion plane can replace the second revolute joint of each leg of a 3-RRR robot, and the resulting kinematics are exactly those of a 3-RPR robot: each leg becomes a base revolute joint plus a prismatic joint of length $\\rho_i$ plus a platform revolute joint. Because the direct and inverse Jacobian matrices factor cleanly for this structure, the determinant of the direct Jacobian, set to zero, yields explicit polynomial singularity surfaces for each of the eight actuation modes (Appendix, Eqs. 31-38). With a symmetrical base and mobile platform, the paper shows that these surfaces are not superposed within the workspace bounded by $r_{\\min} \\le \\rho_i \\le r_{\\max}$, and concludes that any pose can be reached by selecting a non-singular actuation mode. It also derives, from the workspace-boundary equations, the scissor bar lengths and heights needed to guarantee a prescribed regular workspace.","pith_inferences":["If the ideal-scissor substitution survives prototype testing, a three-motor, six-clutch architecture could offer a low-cost way to retrofit planar parallel robots with singularity avoidance, since no extra power amplifiers are needed.","The closed-form singularity equations invite a direct-kinematics and cuspidality analysis per actuation mode, which the paper names as future work; one testable extension is to count direct-kinematics solutions for each mode.","The non-superposition observation is geometric, not yet a proof of global path feasibility; an extension would be to check whether the same property holds under asymmetry of the base and platform or under looser joint limits.","Near the minimum scissor length, physical clearance in the scissor joints may make the serial-singularity limit $\\rho_i = 0$ approachable, so $r_{\\min}$ should be chosen with a safety margin calibrated by measurement."],"forward_implications":["Each of the eight actuation modes has a single inverse-kinematics solution, so switching modes never requires also jumping between different inverse solutions.","The singularity surfaces are available as closed algebraic expressions for the symmetric architecture, which makes exact singularity checks possible for any candidate pose rather than per-orientation numerical sampling.","Because the singularity surfaces do not coincide inside the nominal workspace, the actuation-mode-selection algorithm can route any end-effector path through poses that would be singular in one mode by switching to another mode.","Workspace requirements can be translated directly into scissor dimensions: given the desired regular disc and orientation range, the formulas fix the minimum and maximum leg lengths and hence the bar length, height, and number of scissor stages."],"supporting_citations":[{"why":"Establishes the notion of working modes and multiple inverse-kinematics solutions that this robot's single-solution property is contrasted with.","marker":"[1]"},{"why":"Introduces joint-coupling as a way to manage parallel-manipulator singularities, the conceptual basis for switching actuated joints.","marker":"[5]"},{"why":"Introduces the NaVARo variable-actuation robot and its eight actuation modes, the design this paper evolves.","marker":"[7]"},{"why":"Documents the algebraic difficulty of singularity loci for revolute planar manipulators, motivating the compact equations obtained here.","marker":"[8]"},{"why":"Provides the algebraic tools used to eliminate passive variables and obtain the determinant-based singularity surfaces.","marker":"[15]"},{"why":"Supplies the actuation-mode-selection algorithm reused to avoid singular configurations in the full workspace.","marker":"[18]"}],"fun_headline_variants":["Eight actuation modes let planar robot avoid every singularity","Scissor legs give planar robot eight ways to dodge singularities","Planar robot with scissors: eight singularity-free actuation modes","Scissor mechanism yields eight non-singular actuation modes","Planar parallel robot uses scissor legs to skip singularities"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that each vertically mounted scissor behaves exactly like an ideal sliding joint of length $\\rho_i$ in the robot's plane; if the scissors bend, bind, or constrain the platform out of plane, the Jacobian, singularity surfaces, and workspace bounds do not describe the physical robot.","fun_headline_variants_meta":{"raw":{"variants":["Eight actuation modes let planar robot avoid every singularity","Scissor legs give planar robot eight ways to dodge singularities","Planar robot with scissors: eight singularity-free actuation modes","Scissor mechanism yields eight non-singular actuation modes","Planar parallel robot uses scissor legs to skip singularities"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000927,"raw_usage":{"total_tokens":3982,"prompt_tokens":963,"completion_tokens":3019,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":579,"completion_tokens_details":{"reasoning_tokens":2931}},"tokens_in":579,"tokens_out":3019,"duration_ms":48071,"temperature":1.0,"reasoning_tokens":2931,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T14:54:49.021409+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take the eight determinant equations from the Appendix and search for a real pose $(x, y, \\alpha)$ inside the nominal workspace that satisfies two different mode equations simultaneously; any such pose is a configuration where the claimed strategy cannot find a non-singular actuation mode. The same check could be run on a physical prototype by commanding a path through such a pose and observing whether the platform stiffens or drifts.","supporting_citations":[{"cited_title":"and Wenger, P., Working Modes and Aspects in Fully-Parallel Manipulator,Proceeding IEEE International Conference on Robotics and Automation, pp","cited_arxiv_id":null,"evidence_quote":"Establishes the notion of working modes and multiple inverse-kinematics solutions that this robot's single-solution property is contrasted with."},{"cited_title":"and Li, C., Management of parallel-manipulator singularities using joint-coupling Advanced Robotics, vol","cited_arxiv_id":null,"evidence_quote":"Introduces joint-coupling as a way to manage parallel-manipulator singularities, the conceptual basis for switching actuated joints."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Introduces the NaVARo variable-actuation robot and its eight actuation modes, the design this paper evolves."},{"cited_title":"A., and Gosselin, C","cited_arxiv_id":null,"evidence_quote":"Documents the algebraic difficulty of singularity loci for revolute planar manipulators, motivating the compact equations obtained here."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the algebraic tools used to eliminate passive variables and obtain the determinant-based singularity surfaces."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the actuation-mode-selection algorithm reused to avoid singular configurations in the full workspace."}],"review_version":1}