Blog

Notes on teaching robots new tricks

How the studio works, why tiny robots come first, what the checks catch, and how a policy gets from simulation to a servo.

workflowopen-duck-miniSep 29, 2026 · 6 min readFrom a sentence to a walking duck: how kualia turns a plain-language brief into a training taskA walk through the whole kualia workflow, from a one-sentence brief to an ONNX policy, using the Open Duck Mini and a small ramp.Read the post

Sep 27, 2026 · 5 minWhy we start with tiny robotsDesk-sized robots train on CPU in minutes to hours, cost little to own and cross the sim-to-real gap without heroics. That is why kualia's shelf starts with them.

Sep 25, 2026 · 6 min13 checks before a single training stepalia writes the training task, but it does not get to grade its own work. Every task passes 13 sandbox checks before training can start. Here is what each one catches.

Sep 23, 2026 · 5 minFrom simulation to servo: running your ONNX policy on a real SO-101What is inside a kualia ONNX export, how to read its metadata, a minimal onnxruntime loop for the SO-101, and the sim-to-real checklist before you power the servos.

Sep 21, 2026 · 5 minLet your coding agent train the robotkualia is a remote MCP server. Point Claude Code, Cursor or Codex at it and your agent can pick a robot, build the task, train it and fetch the ONNX policy from your editor.

Research digest

Short reads on new robotics and AI papers, written by our research assistant and reviewed before publishing. 1,060 so far.

Page 27 of 59

Apr 24, 2026Beyond Single Metrics: Why Network Digital Twins Need Mixture of Experts Introduction The proliferation of 5G and 6G networks has introduced a fundamental tension in network management. Emerging applications, from autonomous vehicles requiring ultra-reliable low-latency communication to augm Apr 24, 2026Specialists, Generalists, and the PokéAgent Challenge: Lessons from Competitive PokémonThe release of The PokeAgent Challenge: Competitive and Long-Context Learning at Scale marks a critical inflection point for evaluating frontier AI systems. By establishing standardized benchmarks for competitive Pokémon Apr 24, 2026 Apr 24, 2026 Apr 24, 2026Beyond Word Error Rate: Why Large Language Models Are Reshaping How We Evaluate Speech RecognitionFor decades, the automatic speech recognition community has operated under a convenient fiction. We have optimized models, published leaderboards, and declared victory based on a metric that measures string edit distance Apr 24, 2026Auction-Based Coordination for Dynamic Multi-Objective Reinforcement LearningThe static nature of traditional reinforcement learning stands in stark contrast to the fluidity of real world deployment. Consider a mobile robot tasked with feeding cats in a shelter, where hungry animals appear and di Apr 24, 2026Forward-Looking Active Retrieval: Reimagining Uncertainty as an Epistemic Signal in Language ModelsLarge language models, despite their impressive capacity for language understanding and generation, frequently hallucinate factual content. While retrieval augmented generation has emerged as a primary mitigation strateg Apr 24, 2026Distributed Online Meta Learning: Solving the Cold Start Problem Through Network CollaborationOnline meta learning has emerged as a foundational technique for enabling adaptive intelligence in resource constrained IoT environments. Unlike traditional batch meta learning, which assumes access to large, static data Apr 23, 2026When Can LLMs Learn to Reason with Weak Supervision? A Critical AnalysisThe promise of reinforcement learning with verifiable rewards (RLVR) has captivated the field of large language model research. If models could learn robust reasoning from nothing more than binary correctness signals, we Apr 23, 2026Self-Supervised Learning Through Efference Copies: Unifying Biological Motor Control and Machine LearningThe quest for label efficient learning has propelled self-supervised learning (SSL) to the forefront of machine learning research. Yet beneath the empirical successes of methods like SimCLR, BYOL, and ReLIC lies a fragme Apr 23, 2026The Semantic Gap in Geometric Reasoning: Why LLMs Still Fail at PhD-Level 3D Vision Introduction The promise of AI-assisted research has always hinged on a specific fantasy: a system that reads a technical paper and immediately produces reliable, executable code that captures the author’s mathematical Apr 23, 2026The Architecture of LLM Reasoning in Binary Analysis: Implicit Patterns and Open QuestionsBinary vulnerability analysis has historically relied on static analysis pipelines that construct global program representations in a single pass. The advent of LLM based agents introduces a fundamental shift toward iter Apr 23, 2026Radiometric Consensus: Disentangling Materials and Illumination Through Multi-Object ConstraintsInverse rendering, the recovery of physical scene attributes from images, represents one of computer vision's most persistent challenges. When constrained to a single observation, the problem becomes mathematically under Apr 23, 2026Multimodal AI in Education: When Engagement and Learning DivergeThe rapid integration of AI chatbots into educational settings has created both opportunities and challenges that we are only beginning to understand. A recent study by Taneja, Singh, and Goel titled "Impact of Multimoda Apr 23, 2026Optic Flow Discontinuities as Emergent Attention Mechanisms in Bio-Inspired NavigationFlying insects navigate cluttered environments with remarkable efficiency despite carrying neural circuitry that fits within a cubic millimeter. This disparity between biological capability and computational constraint h Apr 23, 2026 Apr 23, 2026Chain-of-Thought Predictive Control: Reconsidering Unsupervised Subskills in Contact-Rich DomainsImitation learning for low level control remains stubbornly difficult when demonstrations are suboptimal. In contact rich manipulation, human teleoperation or kinesthetic teaching produces trajectories that are inherentl Apr 23, 2026Beyond Architecture Search: Teaching Machines to HypothesizeThe pursuit of optimal neural architectures has long been characterized as an exercise in search. Whether through evolutionary algorithms, reinforcement learning, or differentiable methods, the prevailing metaphor treats