DART-S: Reachability-Audited Active-Suspension Preconditioning for Off-Road Vehicle Jumps

Yu Hu; Fangzhou Zhao; Liang Chen; Chen Min; Wei Li; Mingyuan Sang; Jiajia Ma; Shican Chen

cs.ROAug 20, 2026

DART-S: Reachability-Audited Active-Suspension Preconditioning for Off-Road Vehicle Jumps

Authors: Yu Hu, Fangzhou Zhao, Liang Chen, Chen Min, Wei Li, Mingyuan Sang, Jiajia Ma, Shican Chen,

Organizations: Research Center for Intelligent Computing Systems, Institute of Computing Technology, CAS, Beijing 100190, China. · School of Computer Science and Technology, University of Chinese Academy of Sciences, Beijing 100049, China. · Dong Feng Off-Road Vehicle Co., Ltd, Shiyan 442000, China.

Abstract

Airborne torque reaction cannot recover takeoff errors beyond the wheel angular-momentum budget. DART-S applies ramp-face suspension preconditioning to change pitch, pitch rate, and wheel spin before liftoff, thereby shifting the queried state and altering the remaining authority budget. To predict how each suspension action reshapes this state-budget pair, DART-S employs a local calibration map. A support-aware selector combines the predicted shift with local outcome evidence and an interval-reachability screen; an exact-pair audit reports residual authority. Across 600 new runs in 72 independent BeamNG sessions, every positive, negative, and boundary query follows its prespecified branch. At the confirmed 40°/13 m/s boundary, DART-S attains 24/24 post-touchdown attitude-criterion successes versus 0/24 for DART (session-level Holm-adjusted p=0.0234). At 11.5 m/s, a 0.35 s timing action attains 23/24 versus 0/24 for the static preset (p=0.0156). The 200 rad/s command guard keeps drivetrain hard-limit exceedance at zero across all 600 runs. The source code will be available at https://github.com/MeridianCAS/DART-S

Figures & tables

Fig. 1: DART-S overview. (a) Phase structure: during the approach (P0) the selector latches a per-condition decision (deploy the preset, abort, or remain neutral) early enough that the preset completes before the ramp fillet; DART’s pitch-axis torque-reaction law corrects the airborne phase (P1) and landing metrics latch at first contact (P2). (b) Support-aware selector: calibrated attitude and wheel-spin shifts are evaluated by an operational interval-reachability screen. Local calibration support, outcome evidence, and that screen govern deployment; the exact-pair set reports residual authority diagnostically.

Fig. 2: DART-S platform and test conditions. (a) DART-S full-scale 4WIDS platform; orange overlays identify the series-hydraulic suspension elements and cyan arrows show the body frame. (b) Representative geometry and disturbance classes: calibration, boundary, and washboard tests use tabletop-fillet ramps; convex $R=35/22$ m exit lips are separate mechanism and held-out conditions. Terrain profiles are schematic and not to scale; orange circles and green squares denote takeoff and touchdown.

Fig. 3: Action-minus-neutral landing-pitch difference vs. ramp angle at

v=13

m/s (

N\geq 30

per condition) for the suspension presets and torque-bias baseline. Open markers: not significant (

p\geq 0.05

Fig. 4: Speed-dependent response on the 35 ∘ tabletop fillet (30 pairs per speed). Positive static-pitch authority persists through the shaded 11.5–12 m/s landing-response reversal, motivating timing-conditioned action selection.

test	protocol	measured result and interpretation
static authority	full lift/drop; front-up/rear-down	ride height changes by +8.3 and -6.3 cm; static pitch is +3.5 ∘ /-3.6 ∘
stroke scaling	33/50/67 mm (10/15/20%), $N=30$	$\Delta\theta_{0}$ = +2.5 ∘ /+3.3 ∘ /+4.0 ∘ ; monotonic takeoff authority
takeoff vs both ends	20 ∘ and 35 ∘	+7.0 ∘ versus +7.4 ∘ at 20 ∘ ; +8.6 ∘ versus +13.0 ∘ at 35 ∘
platform robustness	mass $\pm$ 20%; CoG-forward, $N=30$	$\Delta_{\mathrm{land}}$ between +6.5 ∘ and +7.5 ∘ (all $p<10^{-4}$ ), versus +7.4 ∘ nominal
landing-gain stroke scaling	33/50/67 mm (10/15/20%)	$\Delta_{\mathrm{land}}$ = +2.8 ∘ /+7.4 ∘ /+8.2 ∘
landing-end variants	$N=30$	pitch change at most 0.6 ∘ ; peak-deceleration change within $\pm$ 4.8% (not an equivalence test)

TABLE I: Compact validation summary for actuator authority, ablations, robustness, and selector execution.

\Delta_{\mathrm{land}}

( ∘ ) is reported relative to neutral.

Fig. 5: Boundary confirmation across independent BeamNG sessions (post-touchdown attitude-criterion successes out of 24; Wilson 95% intervals). (a) At 40 ∘ /13 m/s, DART-S attains 24/24 versus 0/24 for DART. Daggered conditions have large descriptive effects but fail their prespecified speed-match gates. (b) The 35 ∘ /16 m/s factorial yields a positive interaction estimate, but its session-level sign-flip test is not significant (

p=0.0625

Fig. 6: Mechanism measurements. (a) Paired takeoff pitch-rate difference

\Delta\omega_{\theta 0}

vs. entry speed on the 20 ∘ kicker (calibration

R=35

m, held-out

R=22

m); dashed lines mark

v^{*}=\sqrt{gR\cos\alpha}

; filled markers: sign-test

p<0.05

. (b) Overlaid axle-stroke traces at 18 m/s (

R=35

m): the nose-up preset stretches the lift-off window from 0.04 to 0.36 s (shaded).

Vehicle Dynamics

Vehicle dynamics studies how vehicles move and respond to forces from tires, terrain, aerodynamics, and control inputs.

See all 78 papers

DART: Dual-Axis Airborne Reachability-Gated Torque-Reaction for Off-Road Vehicle Jumps
Jul 31, 2026Yu Hu, Fangzhou Zhao, Mingyuan Sang +7Reachability Analysis Vehicle Dynamics
Research Center for Intelligent Computing Systems, Institute of Computing Technology, CAS, Beijing 100190, China. · School of Computer Science and Technology, University of Chinese Academy of Sciences, Beijing 100049, China. · Dong Feng Off-Road Vehicle Co., Ltd, Shiyan 442000, China.
Saturation-Aware Robust Trajectory Optimization for Reusable Launch Vehicles via Differentiable Physics
Jul 2, 2026Liwei Chen, Tong QinTrajectory Optimization Robust Control
Beijing Institute of Aeronautical Systems Engineering, Beijing 100076, China
Tilt as a Certified Resource: Preserving Motor Wrench-Rate Authority on Articulated Multirotors
Sep 18, 2026Giuseppe Silano, Martin SaskaQuadrotor Control Robot Morphology-Control Co-Design
Department of Power Generation Technologies and Materials, Ricerca sul Sistema Energetico (RSE) S.p.A., 20134 Milan, Italy · Department of Cybernetics, Czech Technical University in Prague, 12135 Prague, Czech Republic

Jul 31, 2026J/K move · Enter open · S save

DART: Dual-Axis Airborne Reachability-Gated Torque-Reaction for Off-Road Vehicle Jumps

Yu Hu, Fangzhou Zhao, Mingyuan Sang, Cheng Min+6

Research Center for Intelligent Computing Systems, Institute of Computing Technology, CAS, Beijing 100190, China. · School of Computer Science and Technology, University of Chinese Academy of Sciences, Beijing 100049, China. · Dong Feng Off-Road Vehicle Co., Ltd, Shiyan 442000, China.

Reachability Analysis Vehicle Dynamics

Open paper AskPDF arXiv

Traversing crests, ledges, and ditches at high speed often launches vehicles into the air, and a mishandled landing presents a substantial crash hazard. We show that the airborne phase is barely controllable: on a 1383 kg platform the wheel angular-momentum budget caps the recoverable pitch-rate change at roughly

9

13^\circ

/s in the tighter nose-up direction under drive at typical takeoff wheel speeds, and at about twice that in the reverse-inclusive braking direction; driving the wheels to their drivetrain hard limit raises the measured nose-up ceiling to only

16

18^\circ

/s. Takeoff pitch-rate disturbances beyond this directional budget are physically unrecoverable in flight, so the decisive leverage lies before takeoff. DART (Dual-Axis Airborne Reachability-Gated Torque-Reaction) back-propagates the landing constraint into a closed-form certified feasible-takeoff set, which supplies a conservative go/no-go condition and a pre-takeoff speed-shaping law. In flight, DART regulates pitch and roll via steer-resolved wheel-reaction torque, governed by a per-flight roll latch derived from the yaw-coupling analysis. In deterministic full-scale simulation in BeamNG.tech, a calibrated pre-takeoff speed regulator reduces touchdown speed by 36% and raises on-target landings from 0/30 to 30/30. Under the same steep-lip approach the airborne law completes 29/30 safe landings under crash-avoidance bounds versus 0/30 for reaction-wheel-style PD (RW-PD) and time-optimal bang-bang (TOBB). On banked run-ups DART holds the median pitch error at or below

2^\circ

at every cross-slope, with the largest baseline separation at

γ=12^\circ

. Across disturbance regimes, the latch preserves pitch-only allocation on low-disturbance entries and enables dual-axis control when roll becomes binding. All results are from simulation; hardware validation remains open.

Abstract

Vehicle Dynamics

Explore similar work

DART: Dual-Axis Airborne Reachability-Gated Torque-Reaction for Off-Road Vehicle Jumps

Saturation-Aware Robust Trajectory Optimization for Reusable Launch Vehicles via Differentiable Physics

Tilt as a Certified Resource: Preserving Motor Wrench-Rate Authority on Articulated Multirotors