.png&w=640&q=75)

Smart Detection Triage is a real-time Physical AI application running locally on the SiMa.ai Modalix DevKit. It continuously processes a camera or RTSP video stream using YOLO26m for object detection, then applies a configurable routing policy based on object class, confidence, and polygon zone. Instead of invoking an expensive vision-language model on every frame, the system selectively sends only important events—such as a high-confidence person detected in a restricted zone—to Qwen3-VL for structured scene enrichment. Vehicles can be suppressed, while unmatched or capacity-limited events are placed into a bounded review queue. The application produces schema-validated JSON alerts containing authoritative detector facts, the matched policy rule, zone, confidence, timestamps, latency measurements, and Qwen-generated visual context. Cooldown, deduplication, queue limits, asynchronous processing, and automatic stream recovery keep continuous detection responsive. A live dashboard allows operators to adjust the confidence threshold without restarting inference and observe FPS, trigger rate, avoided VLM requests, latency, alerts, and review activity. YOLO and Qwen are third-party model families deployed through SiMa-compatible Neat and LLiMa runtimes; the project’s key innovation is the policy-gated orchestration layer that determines when each model earns its computational cost.
16 Sep 2026

nano-agent treats cost as a live reasoning input. At every step, the agent decides: proceed, downgrade to a cheaper model, or skip — governed by a real USDC budget settled on Arc via Circle Gateway. Gemini 2.0 Flash handles fast inference, function calling, and Google Search grounding. Gemini 2.5 Pro verifies factual claims. Each tool call triggers HTTP 402 — the agent signs an EIP-3009 USDC authorization and retries with payment proof. Circle Gateway settles micropayments on Arc in under a second at sub-cent cost. A live dashboard shows spend, earned, step count, and model-tier decisions in real time. Benchmark: 416 steps · $0.240 spent · $0.276 earned · +15% margin · $0.0006/action average. A screen recording shows a $0.25 live run — budget bar draining, model tier switching, decision log printing go · downgrade · skip. Economy view breaks down cost: 77% data, 17% LLM, 4% verification. Two agent wallets earn independent revenue. TAM is $200B+ AI infrastructure. SAM is $15B agentic API economy — every autonomous agent calling external services is a potential payer. Revenue comes from per-step settlement margin, SDK licensing, and premium x402 tool endpoints with built-in payment rails. LangChain, AutoGen, and CrewAI treat cost as post-hoc billing. nano-agent makes it a real-time constraint on behavior. The Arc economics are the moat: the same 417 actions cost roughly $834 in Ethereum gas — 3,475 times more expensive — making per-step micropayments impossible anywhere else. On Arc mainnet, multi-agent economies — agents hiring agents, earning revenue — become self-sustaining. Every x402-compatible API becomes a paid tool layer. nano-agent is the budget runtime that keeps those economies solvent.
26 Apr 2026