Early-Signal Detector

method · first seen 4 Oct 2026

model predictive control

Also: MPC, predictive control, receding-horizon control, receding-horizon planning, receding-horizon motion planning

A control strategy that uses a predictive model to optimize control actions over a receding horizon.

Model predictive control repeatedly solves a finite-horizon optimization over a dynamics model, applies the first action, and re-plans. The sources are routine applications and extensions rather than a new idea: learned-critic MPC for contact-rich tasks, a GPU-accelerated QP solver for linear MPC, game-theoretic MPC for drone coordination, ground-reaction-force-aware MPC for rehabilitation robots, crowd-navigation MPC with Koopman prediction, sampling-based whole-body MPC for humanoids, and VLM-generated cost programs for manipulation. Latent world-model planning papers also use receding-horizon rollouts. Two GitHub repos, a robotics toolkit and a landing-simulation project, also mention it. Mention counts are flat to declining.

MPC has been a standard control method for decades, so this rise reflects normal publication volume. The more specific trends inside it, such as GPU solvers, learned costs or critics, and VLM-specified objectives, are better candidates to track separately.

Analyst note (established), 17 Nov 2026.

Mentions per week

01819 May11 Aug10 NovPapersCodeDiscussion

PapersCodeDiscussion

Rank 181 on 14 Nov 2026 · 49 mentions in 28 days · 47 independent groups · papers → code · first seen 6 weeks ago · acceleration -0.23

Earliest sources

Latest sources

Mentioned together with

control barrier functionflow matchingKoopman operator learningiterative linear quadratic regulatorworld modelGaussian processknowledge distillationimitation learningnonprehensile manipulationwhole-body control