AI & ML
Course-Referenced Lateral Control of Trail-Following Convoys: String Stability and Adaptive Feedforward under Load Uncertainty
arXiv cs.RO0 views
arXiv:2611.20320v1 Announce Type: cross Abstract: Vehicles in a convoy can steer along the GNSS trails of the vehicles ahead when lane markings are unavailable, as on snow covered roads. Since each vehicle's tracking error enters the reference of the next, lateral errors can grow along the convoy. This paper presents a course referenced formulation of trail following in arc length, in which each follower takes the direction of its fitted trail as its heading reference. Error propagation then reduces to a single transfer function. We derive a condition under which combining the trails of the leader and the predecessor keeps the errors bounded for any convoy length, and show that following the predecessor alone lets the errors accumulate without bound, even without amplification, on roads that end turned or shifted sideways. Passenger and cargo loads and tire stiffness vary in practice, and we show that they enter the curvature feedforward through a single coefficient, whose mismatch leaves steady offsets on curves. We propose an adaptive law that estimates this coefficient on curves and removes the offsets. Simulations of an eleven vehicle convoy confirm the analysis.
Source:https://arxiv.org/abs/2611.20320PreviousAerial Person Detection under Edge Constraints: Held-Out Evaluation and Edge-Ground VerificationNextVoxSAOR: Semantic-Aware Occupancy Robustification via Aleatoric Uncertainty Modeling
Related posts

Running Claude Code, Codex and Cursor at the same time? 8 Mac apps that keep three agents from becoming chaos (2026)
0dev.to AI
Google launches Gemini 4 Neon, its Gemini 4 model for coding and knowledge work, at $3/$15 per million tokens
0dev.to AI
Image models can paint a torii. They still invent the 営業時間
0dev.to AI

Gemma 4 on AWS Lambda with a GPU: E2B on an NVIDIA L4, Day One
0dev.to AWS

Keep an AI Coding Handoff Reproducible After One File Changes
0dev.to WebDev

Worse Than AI Cheating: 20 Widely Used AI Models Were Tricked Into Dangerous Actions, and All 20 Went Along
0dev.to Security