{"id":"41071127-f0e6-4c56-b88e-0fe8550ef745","arxiv_id":"2505.02460","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"A 69 kg, 1.17 m automated driving research platform with all-wheel steering and drive, modular ROS architecture, and optimization-based planning and control, demonstrated in two experimental scenarios.","lead":"ZeloS is a small, modular research vehicle for testing automated driving algorithms on real hardware at low cost and risk. This paper describes its design, software architecture, motion planning and control, and shows experiments with tracking errors under 2 cm.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Experimental support for 'reliable navigation and autonomous obstacle avoidance' rests on a single qualitative run and a planning snapshot rather than repeated trials or completion metrics.","rationale":"The reader's verdict of CONDITIONAL is appropriate, and this stress-test pass reinforces rather than overturns it. The paper is a legitimate systems contribution: the hardware is modular, the automation architecture is clearly described, and the optimization-based planning/control stack is standard and coherent. The reader's weakest_assumption focuses on the constant coupling matrix G in Eq. (2), which is a real modeling idealization, but the experimental tracking error suggests it is not the dominant risk for the demonstrated scenario. The more load-bearing concern is the evidence-to-claim gap: 'navigates reliably' and 'autonomously avoids obstacles' are strong conclusions, yet the experiments comprise a single delivery run and a planning snapshot without statistical repetition or obstacle-avoidance completion metrics. This does not change the verdict because the reader already set CONDITIONAL and listed thin experimental evidence as a main limitation; however, it identifies a different primary concern than the stated weakest_assumption. The proposed concrete test would settle whether the reliability claim can be maintained or whether the conclusion must be weakened to a single-run feasibility demonstration.","tokens_in":11906,"tokens_out":4097,"duration_ms":58351,"concrete_test":"Repeat the Section VII delivery scenario at least 10 times under identical conditions and report the per-run maximum and root-mean-square values of the tracking errors epsilon_x, epsilon_y, and epsilon_varphi, along with the success rate (arrival at the target without human intervention). Separately, run at least 5 obstacle-avoidance trials with the obstacle placed at varied positions orthogonal to the reference path, and record the minimum distance between the vehicle body and the obstacle as well as whether the full path was completed. If any trial fails or the maximum lateral error exceeds a pre-specified threshold (e.g., 50 mm), the conclusion should be scoped to a single-run demonstration rather than stated as reliable navigation and autonomous obstacle avoidance.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central conclusion is that the presented experimental results demonstrate that ZeloS navigates reliably from an initial to a target position and autonomously avoids obstacles. The evidence in Section VII, however, consists of one delivery scenario whose tracking error is summarized only as 'at most < 20 mm' (Fig. 11), and one obstacle-avoidance scenario shown as a planning snapshot at t = 2 s (Fig. 12). No repeated trials, no error statistics, no success/failure count, and no obstacle-clearance or completion metric are reported. Figure 12 shows the planned reference trajectory diverging from the reference path, but it does not show whether the vehicle subsequently cleared the obstacle, how close it came, or whether the maneuver was completed. Reliability is a statistical property; a single successful run cannot establish it. The mechanical idealization behind the constant coupling matrix G in Eq. (2) is a plausible modeling risk, but it is partially mitigated by the observed tracking error in the one run. The more load-bearing gap is the mismatch between the thin experimental evidence and the strong conclusion of reliable, autonomous obstacle avoidance.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents ZeloS, a 69 kg, 117 cm research platform for early-stage validation of automated driving methods. It describes the modular hardware (all-wheel steering, all-wheel drive, distributed computation) and a centralized ROS-based automation architecture, then focuses on the implemented motion planner and tracking controller: an MPC-based path-following planner that uses convex polytopic free space from the 2D LIDAR, a cascaded MPC/PI tracking controller, and an analytical force-allocation scheme. The experimental section reports one delivery scenario, for which the tracking error is stated to be below 20 mm, and one obstacle-avoidance scenario shown as a planning snapshot at t = 2 s. The authors conclude that ZeloS navigates reliably and autonomously avoids obstacles.","tokens_in":12126,"tokens_out":7686,"duration_ms":93421,"significance":"If the platform performs as claimed, this is a useful engineering contribution: it provides a moderately sized, modular, sensor-rich testbed for early validation of motion planning and control algorithms, with a digital twin for software-in-the-loop testing. The paper is clearly written, gives a fairly complete hardware and software architecture, and models the vehicle dynamics and control allocation in enough detail to be reproduced. The reported real-world run helps demonstrate that the full stack works end to end. However, the quantitative support for the central claim is currently thin: there is only a single experimental run and a single planning snapshot, with no statistical analysis, repeated trials, or avoidance-completion metrics. The central claim of a functional platform is defensible, but the evidence as presented is not yet sufficient to support the words 'reliably' and 'autonomously avoids obstacles'.","major_comments":[{"comment":"The central conclusion in Section VIII that ZeloS 'navigates reliably' and 'autonomously avoids obstacles' is not supported by the reported experiments. Section VII presents one delivery run, with tracking error summarized in Fig. 11 only as 'at most < 20 mm', and one obstacle-avoidance scenario shown as a planning snapshot at t = 2 s in Fig. 12. There are no repeated trials, no success/failure counts, no error statistics or confidence intervals, and no measure of obstacle clearance or maneuver completion. Reliability is a statistical property; a single run cannot establish it. The obstacle-avoidance figure shows only the planned reference trajectory at one instant, not whether the executed vehicle cleared the obstacle. The authors should either add repeated trials and completion/clearance metrics for both scenarios, or substantially weaken the conclusion to a demonstration of a single successful run and a single planning instance.","section":"Section VII and Section VIII"},{"comment":"Equation (16) as printed applies the rotation matrix only to the reference vector and then subtracts the measured position: xε = R(φhat) [xr yr φr]^T - [xhat yhat φhat]^T. With xr, xhat in the global frame, this expression is not a body-frame error and is not invariant to absolute position; it is also inconsistent with the standard error dynamics in Eq. (17), which are derived from e = R(-φhat)(p_r - p_hat). The correct form should be xε = R(-φhat)( [xr yr]^T - [xhat yhat]^T ) for the translational part. Please correct Eq. (16) or clarify the coordinate definitions that would make it valid; if the implementation uses the standard form, the paper should state so.","section":"Section V-C1, Eq. (16)"},{"comment":"The constant coupling matrix G in Eq. (2) rests on the assumption that tire contact points remain fixed relative to the COG for arbitrary steering angles and suspension deflections, because the steering axes are located exactly above the contact points and the suspension is linear. This assumption is used in odometry (Eq. (12)), force allocation (Eq. (25)), and dynamics (Eq. (4)). The paper offers no direct experimental check of this assumption. Given that the single reported run achieved <20 mm tracking error, the assumption may be acceptable for that scenario, but a targeted validation (e.g., comparing encoder-based odometry against the Vicon ground truth under a steering-angle sweep and under varying suspension loads) would de-risk the model for the broader claim that ZeloS supports validation of various automated driving methods.","section":"Section II-B, Eq. (2)"},{"comment":"The optimization problems that define the planner and the position MPC are not reproducible from the paper because several tuning parameters are omitted: the MPC weight matrices Q, S, R in Eq. (21a), the reinitialization threshold vector xth_epsilon in Eq. (15e), and the polytopic sets V_o and A_o in Eqs. (21c)-(21d). These parameters directly affect the reported computation times and tracking performance. Please report their values or provide a repository or online appendix with the complete implementation.","section":"Section V-B and V-C1, Eqs. (15e) and (21a)"}],"minor_comments":[{"comment":"Equation (11) has a typesetting issue in the bracket structure; the formula should read \\hat v_i = [\\hat\\omega_i r_dyn cos(\\hat\\delta_i), \\hat\\omega_i r_dyn sin(\\hat\\delta_i)]^T.","section":"Section IV-B, Eq. (11)"},{"comment":"The notation G+ is used both for the pseudoinverse of G in Eq. (25) and, in Eq. (12), for the pseudoinverse of G^T. Because these matrices have different dimensions, the notation should be disambiguated, e.g., (G^T)^+ in Eq. (12).","section":"Section IV-B and Section V-C3, Eqs. (12) and (25)"},{"comment":"The caption says 'ZeloS' position is at the green mark' while the green line denotes the reference trajectory; please disambiguate the marker and the line in both the caption and the legend.","section":"Figure 12 caption"},{"comment":"The statement that the steering actuator reaches a desired tire angle 'almost instantly' is not quantified; a rise time or bandwidth would substantiate the modeling assumption in Section III that actuator dynamics can be neglected.","section":"Section II-C"},{"comment":"The PI gains RP and RI are not specified. Since they are part of the implemented velocity control, reporting their values (or referencing accessible code) would improve reproducibility.","section":"Section V-C2, Eq. (23)"}],"recommendation":"major_revision","confidential_remarks":"This is a systems and platform paper whose main value is the complete, modular, real-world testbed rather than a new algorithmic idea. The reported single-run validation is too thin for the paper's strong concluding claims, but the missing evidence can be added within the scope of the paper. There is also a likely typo in Eq. (16) that should be fixed before acceptance. If the authors add repeated trials, obstacle-clearance/completion metrics, and the omitted tuning parameters, the paper would become a solid engineering contribution to the autonomous-driving research-platform literature."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things you should know. First, ZeloS is a genuinely useful engineering contribution: a 69 kg all-wheel steering/drive platform with a modular hardware and software architecture, a sensible safety concept, and an optimization-based planning and control stack. The description is concrete and the design choices are well motivated. Second, the experimental section is much thinner than the conclusion. The claim that ZeloS 'navigates reliably' from start to goal and 'autonomously avoids obstacles' rests on a single delivery run (tracking error reported as 'at most < 20 mm' with no statistics or repeated trials) and one planning snapshot at t = 2 s that shows the planned trajectory diverging around an obstacle but does not show whether the vehicle actually cleared it. That mismatch is the load-bearing soft spot. What is new: the platform itself, as an integrated configuration. The individual components - MPC, PI velocity control, analytical control allocation, LIDAR SLAM, digital twin - are established techniques, and the paper honestly cites prior work, including the authors' own [16]. There is no new algorithmic content, and there does not need to be; the contribution is the integration and the public description of the platform. The mathematical modeling is standard and internally consistent. The fixed-tire-contact assumption is a reasonable idealization for a platform designed for simple modeling, and the one run's small tracking error gives some comfort, though it does not fully test the assumption. The soft spots, in proportion: the experimental evidence is qualitative and underpowered for the word 'reliable.' Omitted implementation parameters (MPC weights, thresholds, polytope definitions) make reproduction harder. No code or data release is mentioned. These are fixable in revision but should not be ignored. The self-citation to the group's velocity control is not circular, because the platform is validated experimentally, so I would not hold that against them. Bottom line: this is a legitimate systems paper that deserves a serious referee and would likely be accepted after strengthening the experimental section. Who is it for? Anyone building or using small-scale automated driving testbeds, especially for early-stage validation of planning and control. I would bring it to a reading group if the conversation is about research infrastructure, and I would cite it if I were building a similar platform - but I would want more data and error bars first. My recommendation: send it to peer review, but tell the authors to either soften the reliability claim or back it with a few repeated runs and a clear metric.","headline":"A well-engineered modular research platform whose central validation claim outruns the experimental evidence: one qualitative run supports functioning, not reliability.","tokens_in":778,"tokens_out":1576,"would_cite":false,"duration_ms":28581,"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":"This paper presents ZeloS, a 69-kg modular research platform for early-stage validation of automated driving methods, and shows experiments in which it navigates to a target and avoids obstacles.","keywords":["automated driving research platform","modular automation architecture","motion planning","model predictive control","control allocation","all-wheel steering","localization","digital twin"],"falsifier":"Measure the tire contact point positions relative to the COG while sweeping steering angles and deflecting the suspension, or run the vehicle through aggressive maneuvers with tire slip and compare the encoder-based odometry (Eq. 12) against an external motion capture ground truth. If the odometry error grows in a way the constant-$G$ model cannot explain, or if the force allocation produces unintended yaw moments, the central assumption is falsified.","tokens_in":1611,"feed_emoji":"🤖","tokens_out":4612,"duration_ms":109792,"temperature":0.7,"pith_summary":"This paper introduces ZeloS, a 69-kg, 117-cm-long modular vehicle built as a general-purpose research platform for testing automated driving methods early in development, rather than as a purpose-built demonstrator. The authors claim that modular hardware, a modular automation architecture, and a deliberately simple chassis model make it easy to swap individual autonomy components without deep onboarding. The implemented baseline uses optimization-based motion planning and control: a model predictive planner linearizes the vehicle dynamics, constrains trajectories to a convex collision-free region extracted from LIDAR point clouds, and a tracking MPC keeps the vehicle on the trajectory. Experimental results show ZeloS navigating reliably from a start to a goal position and autonomously avoiding obstacles, with position tracking errors below 20 mm.","feed_headline":"69-kg ZeloS robot navigates and dodges obstacles in tests","feed_subtitle":"Modular hardware and convex model-predictive control on a 69-kg testbed for early autonomy validation.","key_machinery":"The central object is the constant coupling matrix $G$ (Eq. 2): because the steering axes are located exactly above the tire contact points and the suspension is linear, the contact points stay fixed relative to the center of gravity, so one constant $3 \\times 8$ matrix maps tire forces to the COG force and torque vector, and its transpose maps COG velocity to the eight tire velocities. The Moore-Penrose inverse $G^+$ of $G$ carries both the encoder-based odometry (Eq. 12) and the analytical control allocation (Eq. 25), so the platform's model, localization, and control all rest on this single geometric idealization. Around that, the automation uses a convex model predictive planner that linearizes the dynamics along the previous solution and restricts the trajectory to a convex collision-free polytope built from LIDAR point clouds, plus a 100 Hz tracking MPC with feedforward velocity and acceleration and a PI velocity controller.","core_discovery":"The paper's central claim is that a deliberately simple, modular vehicle design can serve as a versatile research platform for validating automated driving methods at an early stage. Each of the four drive modules carries its own actuator, power, and computer, so the wheel arrangement and steering mode can be reconfigured, and the modular automation lets individual modules such as sensors, perception, planning, and control be swapped. The implemented baseline uses a constant coupling matrix G derived from the chassis geometry to relate tire forces and velocities to the center of gravity, enabling a compact dynamic model. Motion planning is a model predictive planner that linearizes the dynamics and constrains the trajectory to a convex, collision-free polytope extracted from LIDAR point clouds, and tracking control is a 100 Hz MPC cascaded with a PI velocity controller and analytical control allocation. Experimental results in the paper show position tracking errors below 20 mm, planner and controller cycle times meeting their targets, and a scenario where the platform avoids an obstacle and navigates from start to goal.","pith_inferences":["A natural extension is to quantify when the constant-coupling assumption degrades: under high tire slip or suspension deflection, odometry and force allocation will be biased, and comparing encoder odometry against an external motion capture system during aggressive maneuvers would reveal the operating envelope.","The modular architecture suggests ZeloS could serve as a shared benchmark platform on which different research groups swap planning or control modules and compare methods on identical hardware.","Given the paper's stated motivation of validating formally safe driving functions, the small scale and low risk make ZeloS a plausible testbed for safety-certified planning and control before full-scale vehicle tests.","A testable extension is to use the digital twin to inject tire slip and suspension deflection, then check whether the tracking error and force allocation remain acceptable compared with the real vehicle, isolating the effect of the constant-coupling idealization."],"forward_implications":["New motion planning or control modules can be validated on ZeloS by replacing only the relevant automation module, because the interfaces are modular and self-contained.","The convex optimization formulation with linearized dynamics runs in real time on the onboard computer, with average planner cycle time 37.54 ms and MPC cycle time 5.43 ms, and more than 99.7% of MPC cycles met the target cycle time.","Position tracking error stayed below 20 mm in the presented scenario, indicating that the platform can support demanding control validations.","The obstacle-avoidance experiment shows that the planner can keep the reference trajectory inside a convex, collision-free admissible set computed from LIDAR data.","The digital twin provides software-in-the-loop testing with the same interfaces as the real vehicle, allowing fast and risk-free checks before hardware experiments."],"supporting_citations":[{"why":"Supplies the simplified tire model, the constant coupling matrix G, the analytical control allocation, and the force-to-actuator inversion used throughout.","marker":"[12]"},{"why":"Provides the differentially flat velocity feedforward control used in the velocity control cascade.","marker":"[16]"},{"why":"Supplies the SLAM algorithm that builds the occupancy map and corrects odometry drift.","marker":"[13]"},{"why":"Provides the method for extracting large convex polytopes from LIDAR point clouds, used to define the collision-free admissible set.","marker":"[14]"},{"why":"Supplies the bi-level planning-and-control architecture that coordinates the motion planner and the tracking controller.","marker":"[15]"},{"why":"Solves the nonlinear motion-planning optimization problem in real time.","marker":"[17]"},{"why":"Solves the position MPC quadratic program at the 100 Hz control rate.","marker":"[18]"},{"why":"Performs algorithmic differentiation and time discretization for the planning and control optimizations.","marker":"[19]"}],"fun_headline_variants":["ZeloS: 69-kg modular robot for early self-driving tests","Modular ZeloS vehicle accelerates early autonomous driving research","ZeloS robot tracks 20-mm accuracy in autonomous validation runs","ZeloS: modular testbed for early-stage autonomous driving validation","ZeloS navigates, dodges obstacles in early self-driving validation"],"cache_read_input_tokens":14848,"weakest_assumption_plain":"The load-bearing premise is that the tire contact points remain fixed relative to the center of gravity because the steering axes sit exactly above the contact points and the suspension is linear; if this geometric idealization fails, the constant coupling matrix and the odometry and force allocation built on it become biased.","fun_headline_variants_meta":{"raw":{"variants":["ZeloS: 69-kg modular robot for early self-driving tests","Modular ZeloS vehicle accelerates early autonomous driving research","ZeloS robot tracks 20-mm accuracy in autonomous validation runs","ZeloS: modular testbed for early-stage autonomous driving validation","ZeloS navigates, dodges obstacles in early self-driving validation"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000728,"raw_usage":{"total_tokens":3246,"prompt_tokens":914,"completion_tokens":2332,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":530,"completion_tokens_details":{"reasoning_tokens":2240}},"tokens_in":530,"tokens_out":2332,"duration_ms":16716,"temperature":1.0,"reasoning_tokens":2240,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T00:49:52.645002+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the tire contact point positions relative to the COG while sweeping steering angles and deflecting the suspension, or run the vehicle through aggressive maneuvers with tire slip and compare the encoder-based odometry (Eq. 12) against an external motion capture ground truth. If the odometry error grows in a way the constant-$G$ model cannot explain, or if the force allocation produces unintended yaw moments, the central assumption is falsified.","supporting_citations":[{"cited_title":"Moseberg, Regelung der Horizontalbewegung eines ¨uberak- tuierten Fahrzeugs unter Ber ¨ucksichtigung von Realisierungsan- forderungen","cited_arxiv_id":null,"evidence_quote":"Supplies the simplified tire model, the constant coupling matrix G, the analytical control allocation, and the force-to-actuator inversion used throughout."},{"cited_title":"Robust position and velocity tracking control of a four-wheel drive and four-wheel steered electric vehicle,","cited_arxiv_id":null,"evidence_quote":"Provides the differentially flat velocity feedforward control used in the velocity control cascade."},{"cited_title":"Real-time loop closure in 2d lidar slam,","cited_arxiv_id":null,"evidence_quote":"Supplies the SLAM algorithm that builds the occupancy map and corrects odometry drift."},{"cited_title":"Werling, Ein Neues Konzept f ¨ur die Trajektoriengenerierung und -Stabilisierung in Zeitkritischen Verkehrsszenarien","cited_arxiv_id":null,"evidence_quote":"Supplies the bi-level planning-and-control architecture that coordinates the motion planner and the tracking controller."},{"cited_title":"On the implementation of an interior-point filter line-search algorithm for large-scale nonlinear programming,","cited_arxiv_id":null,"evidence_quote":"Solves the nonlinear motion-planning optimization problem in real time."},{"cited_title":"HPIPM: A high-performance quadratic programming framework for model predictive control *,","cited_arxiv_id":null,"evidence_quote":"Solves the position MPC quadratic program at the 100 Hz control rate."},{"cited_title":"CasADi: A software framework for nonlinear optimization and optimal con- trol,","cited_arxiv_id":null,"evidence_quote":"Performs algorithmic differentiation and time discretization for the planning and control optimizations."}],"review_version":1}