{"id":"f4c3b81c-249d-42f9-87c4-0bd6cd15b8a2","arxiv_id":"2411.13770","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"A passive shoulder exoskeleton with a pulley-adjustable peak assistive torque angle reduced shoulder muscle activation by up to 49.6% during overhead screwing while preserving range of motion.","lead":"Researchers built a passive shoulder exoskeleton whose peak assistance angle can be adjusted by moving pulleys, so the same device can assist overhead tasks at different arm heights. In a screwing test with ten people, the device cut muscle activity by up to 49.6% when tuned to the task and did not restrict shoulder range of motion.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"PATA adjustment is never validated by direct torque measurement, so the Match/Mismatch EMG differences may not isolate the effect of peak torque angle alignment.","rationale":"The most load-bearing assumption is that the adjustable PATA mechanism behaves as modeled, because the experimental design depends on it to create configurations that differ only in the timing of peak assistance. The reader's weakest_assumption identifies exactly this: the ideal geometric model with an un-derived k and no direct torque measurement. I agree with that assessment. The paper has independent support for the device's general effectiveness—ROM preserved (Section IV-A) and consistent EMG reductions when wearing the exoskeleton—but those observations do not require the PATA mechanism to work as claimed. The specific novelty of the paper is the adjustability of PATA, and that is not validated by any direct measurement. If a torque bench test confirms that the two Match configurations have equal peak torque and PATA values matching the intended task angles, the interpretation is sound; if not, the EMG differences could be attributed to other changes in the torque profile. The sample size, all-male cohort, and laboratory setting are acknowledged limitations but are not the central logical weak point. Therefore the verdict remains CONDITIONAL, consistent with the reader's conclusion.","tokens_in":15629,"tokens_out":4938,"duration_ms":85085,"concrete_test":"Mount the HIT-POSE on a test rig with a torque sensor at the sagittal F/E joint axis. For the two configurations used in Session 2 (Match at low and high heights, with the same spring pretension), measure the assistive torque over the full sagittal F/E range (e.g., 0° to 170°). From the measured torque-angle curves, determine (1) the peak torque magnitude and (2) the PATA (angle at peak torque). Verify that the peak torque in the two configurations differs by less than 5%, and that the measured PATA values are within ±5° of the intended 90° and 120°. Additionally, sweep β over its full range (−30° to 30°) and fit Eq. (4) to obtain an empirical k. If the measured PATA deviates by more than 5° from the model prediction, or if the peak torque is not matched, the Match/Mismatch comparison cannot be interpreted as isolating the PATA effect.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central novelty is the adjustable peak assistive torque angle (PATA), but the actual torque-angle relationship produced by the mechanism is never measured. Section II-C presents the model: Eqs. (1)-(3) assume an ideal cable-pulley-spring system with no friction or cable deformation, and Eq. (4) gives PATA = θ0 + α − kβ, where k is only described as 'a variable positive coefficient around 1.' No direct torque sensor or load-cell data are reported anywhere. In Session 2 (Section III-D), the 'Match' and 'Mismatch' configurations are selected from this model under the assertion that they have equal peak torque and differ only in PATA. If the real mechanism deviates from the model—for example, if moving pulleys 3 and 4 also changes the effective peak torque, or if the true PATA differs from the intended 90°/120° because k is not exactly 1—then the significant EMG reductions in the Match condition (Table VI) and the perceived-effort improvements could reflect unintended differences in torque magnitude or profile shape rather than proper PATA alignment. The device evidently provides assistance, and one configuration outperforms another, but the paper's distinct contribution—adjustable PATA—is not directly verified, leaving the central interpretation unsupported.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes a passive occupational shoulder exoskeleton (HIT-POSE) with an adjustable peak assistive torque angle (PATA) for overhead tasks. The design combines an ergonomic shoulder structure selected via kinematic simulation and a cable-pulley-spring torque generator whose PATA can be changed by moving pulleys along a guide rail. The authors report ROM measurements showing no significant restriction relative to no exoskeleton, and a 10-participant experimental study of a screwing task in which muscle activation, perceived effort, and frustration were reduced in the 'Match' condition (PATA aligned with the task angle) compared with both 'Mismatch' and no-exoskeleton conditions. The central claim is that the adjustable PATA enables efficient assistance across different overhead task angles.","tokens_in":15926,"tokens_out":2820,"duration_ms":28823,"significance":"If the central claim holds, the HIT-POSE would represent a useful step toward task-adaptable passive exoskeletons, addressing a recognized limitation of fixed torque-profile devices. The paper's strengths include direct ROM measurements with a motion-capture system, a randomized within-subject EMG protocol with eight muscles, and the inclusion of both objective and subjective outcomes. The ROM verification and the observation that Match conditions yield greater EMG reductions than Mismatch conditions are valuable empirical contributions. However, the load-bearing validation of the adjustable-PATA mechanism is currently incomplete: the torque-angle relationship is only modeled, not measured, and the experimental conditions are defined from that unverified model. The significance of the EMG results therefore depends on assumptions about the physical device that the manuscript does not yet establish.","major_comments":[{"comment":"The paper states that the Match and Mismatch configurations 'present the same peak torque but different PATAs,' yet Table IV indicates that β, the parameter used to change PATA, has a positive effect on peak torque in Phase II. The manuscript does not explain how peak torque was held constant while β was varied (e.g., whether Li or spring pretension was adjusted to compensate). Without this information, the observed EMG differences between Match and Mismatch conditions could be confounded by differences in torque amplitude rather than by PATA alignment. Please report the actual torque profiles of the two configurations or describe the compensation procedure in detail.","section":"III-D and Table IV"},{"comment":"The PATA formula θ_PATA = θ0 + α - kβ relies on a coefficient k that is described only as 'a variable positive coefficient around 1' with no derivation or experimental identification. Since the Match and Mismatch conditions are defined by setting PATA to 90° and 120° from this formula, the validity of the entire experimental contrast rests on an unverified model parameter. A direct measurement of the assistive torque versus shoulder angle curve is needed to confirm that the physical device actually achieves the intended PATAs and that the two configurations differ only in PATA.","section":"II-C, Eq. (4)"},{"comment":"The torque profile simulations explicitly assume an ideal situation with no cable deformation and no friction. In a real cable-pulley-spring mechanism, friction and cable compliance can shift the peak torque angle and alter the torque magnitude. The manuscript does not provide any bench-top torque measurement to quantify these deviations. Given that the central contribution is the adjustable PATA and its effect on assistance, the absence of any physical torque-angle validation leaves the central interpretation unsupported. I recommend adding a static or quasi-static torque measurement of the device for at least the two experimental configurations.","section":"II-C and Fig. 9"}],"minor_comments":[{"comment":"The text refers to 'Table ??' when discussing ROM results; this should be Table V.","section":"V-A"},{"comment":"In the first paragraph, 'PATA adjus module' appears to be a typo for 'PATA adjustment module.'","section":"II-C"},{"comment":"The abbreviation 'RPD' appears in Section V-D ('RPE, RPD, and PW') but is not defined earlier; the paper only defines RPE and the NASA-TLX subscales.","section":"III-D"},{"comment":"The hex key mass is given as 8.6 g, which seems quite light for an overhead screwing task; please verify that this value is correct or clarify whether the mass refers to the tool only.","section":"III-C"},{"comment":"For the pairwise EMG comparisons, only significance stars are shown; reporting test statistics or exact p-values would improve reproducibility and allow readers to assess effect sizes.","section":"IV-B and Fig. 13"},{"comment":"The axes of Figure 8 are not labeled in the caption; adding axis names (e.g., β in degrees and PATA in degrees) would make the figure self-contained.","section":"Fig. 8"}],"recommendation":"major_revision","confidential_remarks":"The manuscript addresses a relevant problem and the experimental data are plausible, but the missing torque measurement is a standard expectation for exoskeleton papers claiming a specific torque-profile attribute such as PATA. The review process should ask for direct torque validation or, failing that, a substantial re-framing of the claims to acknowledge that the PATA effect is inferred rather than directly verified."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The HIT-POSE paper is worth a look if you work on passive shoulder exoskeletons. The genuinely new piece is the adjustable peak assistive torque angle (PATA) achieved by moving pulleys 3 and 4 along a guide rail. Prior systems like H-PULSE and MATE tune assistance amplitude but not the phase of the peak torque, so this is a real design contribution. The shoulder structure also seems well thought out: measured ROM with the device is close to free movement (164° vs 165° sagittal, 158° vs 160° horizontal), and the EMG reductions are substantial, especially in the Match condition, with up to roughly 50% relative reduction in some muscles. The subjective data (RPE, TLX, SUS) are consistent with the objective findings. That is decent engineering and honest reporting.\n\nThe soft spot is exactly where the stress-test note lands. The torque-angle relationship is only simulated, never measured. Equation (4) predicts PATA with a coefficient k that is 'around 1' and not derived from the mechanics. The paper asserts that the Match and Mismatch configurations have the same peak torque and differ only in PATA, but there is no load-cell or torque sensor data to back that up. If moving the pulleys also changes the torque magnitude or shifts the profile in a different way than the model predicts, the EMG differences between Match and Mismatch might reflect torque amplitude, not PATA alignment. That does not destroy the paper—the device clearly provides assistance, and the Match condition helps more—but it does mean the central novelty is not directly verified. A simple bench test measuring torque at multiple angles would settle it.\n\nSmaller issues: ten male participants, all laboratory-based, no field data, and no mention of corrections for multiple comparisons across eight muscles. These are minor and mostly acknowledged in the limitations.\n\nWho is this for? Researchers working on torque profile generation in passive exoskeletons, and anyone comparing adaptability mechanisms. It deserves a serious referee because the design idea is plausible and the experimental protocol is more careful than many hardware papers. My recommendation: send it to review, but ask the authors for a direct torque measurement of the adjustable-PATA mechanism and a clearer statistical handling of the multiple muscle comparisons. Conditional accept feels about right.","headline":"Solid hardware paper with a real novelty in adjustable torque phase, but the central PATA claim is only simulated, not measured, so the Match/Mismatch result is weaker than it looks.","tokens_in":16425,"tokens_out":1950,"would_cite":true,"duration_ms":20753,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"A passive shoulder exoskeleton with an adjustable peak assistive torque angle cuts muscle activation by up to 49.6% while preserving range of motion.","keywords":["passive shoulder exoskeleton","overhead work","peak assistive torque angle","muscle activation","shoulder range of motion","kinematic compatibility","torque generator","work-related musculoskeletal disorders"],"falsifier":"Mount the exoskeleton on a test rig, measure the output torque at the shoulder with a load cell while sweeping the sagittal angle, and compare the measured peak angle to $\\theta_{\\mathrm{PATA}} = \\theta_0 + \\alpha - k\\beta$ across the advertised pulley travel; a systematic offset larger than the angle difference between the Match and Mismatch conditions would collapse the explanation of the EMG differences.","tokens_in":15424,"feed_emoji":"🦾","tokens_out":7118,"duration_ms":66324,"temperature":0.7,"pith_summary":"Overhead work strains the shoulder, and passive exoskeletons that compensate arm weight are a practical industrial fix, but most are built for one task with a fixed torque profile that peaks at one arm angle. This paper proposes a passive shoulder exoskeleton, HIT-POSE, whose peak assistive torque angle can be adjusted by moving pulleys along a guide rail, so the same device can be reconfigured for tasks at different shoulder elevations without being rebuilt. In ten participants doing a screwing task at two heights, the exoskeleton matched to the task reduced muscle activation in all eight monitored muscles (relative reduction up to 49.6%) and lowered perceived effort and frustration, while measured shoulder range of motion stayed statistically unchanged. The significance, if the result holds, is that one lightweight passive exoskeleton could adapt across overhead tasks and individual body sizes, making the assistive technology more usable and effective in real workplaces.","feed_headline":"Torque-angle dial cuts exoskeleton muscle effort by 49.6%","feed_subtitle":"A passive shoulder exoskeleton tunes the timing of peak help to match the task angle.","key_machinery":"The enabling mechanism is a cable-and-pulley torque generator with a passive clutch: a wire rope connects a parallel spring group to the sagittal F/E joint bar, storing elastic energy as the arm lowers and releasing it as the arm raises. The assistive torque is $\\tau_{\\mathrm{exo}} = F_s L_{\\mathrm{exo}}$, with spring force $F_s = K\\Delta L$ and moment arm $L_{\\mathrm{exo}}$. Because the peak of the torque profile is dominated by the moment arm's geometry, moving pulleys 3 and 4 along a guide rail changes angle $\\beta$ and shifts the peak assistive torque angle according to $\\theta_{\\mathrm{PATA}} = \\theta_0 + \\alpha - k\\beta$ (with $\\theta_0 \\approx 80°$ and $k \\approx 1$), while $\\alpha$ sets the critical angle between the low- and high-assistance phases. The shoulder structure itself is a two-degree-of-freedom linkage whose design parameters ($\\phi$, $d_v$, $d_b$) were chosen from kinematic simulation to keep the exoskeleton compact while avoiding collision at high elevation angles, preserving the user's range of motion.","core_discovery":"The central claim is that the timing of peak assistive torque is a tunable, task-relevant degree of freedom for passive shoulder exoskeletons, and that tuning it correctly is what makes assistance efficient. The HIT-POSE generates torque with a parallel spring group routed through pulleys to a sagittal flexion/extension joint bar, and the peak angle is shifted by changing the angle β of a movable pulley pair, following θ_PATA = θ0 + α − kβ. In the experimental screwing task, when the set peak assistive torque angle matched the task's target shoulder angle, all eight measured muscles showed significantly reduced activation (absolute reductions 15.6% to 23.8%, relative reductions up to 49.6%), roughly double the reduction seen in the mismatched configuration. The same match conditions also produced significantly lower perceived exertion, physical demand, effort, and frustration, with a System Usability Scale score of 79.7. The authors read these results as evidence that the exoskeleton can provide natural range of motion and efficient, adaptable assistance for overhead work, thereby reducing the risk of work-related musculoskeletal disorders.","pith_inferences":["Beyond the paper, the strongest test of the thesis would be a bench test that directly measures the exoskeleton's torque-angle curve; if real friction, cable stretch, or pulley behavior shifts the measured peak angle, the advertised 90° to 150° PATA range may be narrower in practice.","The PATA formula and parameter sensitivity analysis suggest that individualized assistance could be set from simple anthropometric measurements such as arm length and target elevation, enabling a lookup-table style tuning procedure for workers of different heights without iterative trial-and-error.","The passive-clutch concept could generalize to other joints: any task with a dominant gravitational load and a well-defined peak torque angle, such as hip or lumbar support during lifting, could use the same movable-pulley phase-shift mechanism.","Because the mismatched configuration still reduced activation in most muscles but not in pectoralis major or latissimus dorsi, a poorly timed torque profile may spare some muscles while leaving stabilizers unevenly loaded; future studies could test whether mismatch profiles create uneven muscle loading."],"forward_implications":["If the PATA-adjustment mechanism works as modeled, one passive exoskeleton can cover overhead tasks at different arm angles by moving pulleys along a guide rail, instead of rebuilding the device for each task.","In the match condition, muscle activation dropped by at least about 15% absolutely and 30% relatively across all eight monitored muscles, and match reductions were about double the mismatch reductions, indicating that alignment of peak torque timing, not merely wearing the device, drives the benefit.","Measured maximum sagittal and horizontal shoulder flexion/extension angles with the exoskeleton (164.46° and 158.28°) were statistically indistinguishable from bare-shoulder values, so the compact shoulder structure does not restrict range of motion.","Perceived exertion and the physical-demand, effort, and frustration components of perceived workload improved significantly in the match condition, and the 79.7 System Usability Scale score supports acceptance in industrial settings.","The reduced muscle activation, including in the erector spinae, implies lower biomechanical load on the shoulder girdle and back, and therefore the potential to reduce the incidence of work-related musculoskeletal disorders with long-term use."],"supporting_citations":[{"why":"Supplies the prior compact passive exoskeleton with tilted and offset shoulder joints whose limited horizontal flexion/extension motivates the ergonomic shoulder structure.","marker":"[11]"},{"why":"Provides the semi-passive exoskeleton with motorized assistance tuning that serves as the comparison for adjustable assistance and for reductions scaling with support level.","marker":"[12]"},{"why":"Documents an overhead-task passive exoskeleton with fixed peak torque timing, cited as evidence that existing devices cannot adapt PATA to different tasks.","marker":"[22]"},{"why":"Frames the kinematic compatibility and adaptability issues in passive cable-driven occupational shoulder exoskeletons that the paper targets.","marker":"[23]"},{"why":"Supplies the Proto-MATE exoskeleton with adjustable peak-torque knob but no PATA adjustment, used as the prior-art comparison for adjustable assistance.","marker":"[25]"},{"why":"Provides field and laboratory evidence that passive occupational shoulder exoskeletons reduce muscle activity and fatigue during overhead work, supporting the outcome interpretation.","marker":"[26]"},{"why":"Supports the claim that individuals of different heights require different peak assistive torque angles, motivating personalization.","marker":"[28]"},{"why":"Provides the surface EMG electrode placement guidelines that the eight-muscle activation measurements rely on.","marker":"[33]"}],"fun_headline_variants":["Tune the assist peak, slash shoulder effort 49.6%","Angle-matched torque peak cuts muscle effort nearly in half","Dial torque peak to task: shoulder effort drops 49.6%","Exo torque timing cuts shoulder effort 49.6%","Adjustable peak assist torque angle reduces effort 49.6%"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The paper assumes the actual torque-angle curve matches the ideal geometric model of Section II-C, with no friction, no cable deformation, and coefficient k near 1, so the adjusted peak angle really lands at the intended arm angle; no direct torque measurement is reported.","fun_headline_variants_meta":{"raw":{"variants":["Tune the assist peak, slash shoulder effort 49.6%","Angle-matched torque peak cuts muscle effort nearly in half","Dial torque peak to task: shoulder effort drops 49.6%","Exo torque timing cuts shoulder effort 49.6%","Adjustable peak assist torque angle reduces effort 49.6%"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001464,"raw_usage":{"total_tokens":5951,"prompt_tokens":1066,"completion_tokens":4885,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":682,"completion_tokens_details":{"reasoning_tokens":4796}},"tokens_in":682,"tokens_out":4885,"duration_ms":32332,"temperature":1.0,"reasoning_tokens":4796,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T15:53:41.407164+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Mount the exoskeleton on a test rig, measure the output torque at the shoulder with a load cell while sweeping the sagittal angle, and compare the measured peak angle to $\\theta_{\\mathrm{PATA}} = \\theta_0 + \\alpha - k\\beta$ across the advertised pulley travel; a systematic offset larger than the angle difference between the Match and Mismatch conditions would collapse the explanation of the EMG differences.","supporting_citations":[{"cited_title":"Design and experimental evaluation of a semi-passive upper-limb exoskeleton for workers with motorized tuning of assistance,","cited_arxiv_id":null,"evidence_quote":"Provides the semi-passive exoskeleton with motorized assistance tuning that serves as the comparison for adjustable assistance and for reductions scaling with support level."},{"cited_title":"A light-weight passive upper arm assistive exoskeleton based on multi-linkage spring-energy dissipation mechanism for over- head tasks,","cited_arxiv_id":null,"evidence_quote":"Documents an overhead-task passive exoskeleton with fixed peak torque timing, cited as evidence that existing devices cannot adapt PATA to different tasks."},{"cited_title":"Design and evaluation of a passive cable-driven occupational shoulder exoskeleton,","cited_arxiv_id":null,"evidence_quote":"Frames the kinematic compatibility and adaptability issues in passive cable-driven occupational shoulder exoskeletons that the paper targets."},{"cited_title":"An experimental evaluation of the proto-mate: a novel ergonomic upper-limb exoskeleton to reduce workers’ physical strain,","cited_arxiv_id":null,"evidence_quote":"Supplies the Proto-MATE exoskeleton with adjustable peak-torque knob but no PATA adjustment, used as the prior-art comparison for adjustable assistance."},{"cited_title":"Analysis and preliminary design of a passive upper limb exoskeleton,","cited_arxiv_id":null,"evidence_quote":"Supports the claim that individuals of different heights require different peak assistive torque angles, motivating personalization."}],"review_version":1}