USDA’s Digital Agricultural Systems Hub (DASH) is combining computer vision, AI, and sensor data from satellites, drones, robots, and farm equipment to advance precision agriculture. Its 50+ TB National Agricultural Image Repository supports model training, while BenchBot and ModCam enable automated crop imaging, weed mapping, disease detection, and precision nitrogen management at field scale. [link]

Volumental and Ochy have integrated 3D foot scanning with AI-powered gait analysis, combining foot shape and running biomechanics into one shoe recommendation. Now live across Runners Need stores in the UK, the system uses Volumental’s scanner and Ochy’s smartphone-based computer vision on one tablet. The companies plan wider retail expansion, showcasing the integration at ISPO Amsterdam. [link]

Bengaluru-based physical AI company SwitchOn is expanding into Germany with a new Berlin location, targeting manufacturers with its DeepInspect® automated visual inspection platform. The system reportedly delivers 99.5% defect-detection accuracy at speeds exceeding 1,200 products/minute, reducing quality costs by up to 50%. Already deployed across 170+ production lines, SwitchOn raised $8M in pre-Series B funding. [link]

Quartz Claims has integrated RAVIN.ai’s AI-powered vehicle inspection into its Australian motor claims workflow. Using smartphone cameras and an app-free browser experience, RAVIN analyzes vehicle damage at FNOL, enabling instant severity assessment, early total-loss identification, repair recommendations, and automated reporting. The partnership aims to reduce claim turnaround from days to minutes while improving objectivity. [link]

Nuance Labs raised $50M Series A, led by Lightspeed with Accel, South Park Commons, NVIDIA and Define Ventures. Its full-duplex AI model understands words, tone, gaze and gestures, responding naturally in real time. [link]

Shanghai-based Qubot Technology raised nearly RMB100M in Pre-A+ funding to develop its AI-powered magnetic-control surgical robot. Its VLA model, 0.1mm catheter-tip control and 4D-DSA digital twin aim to automate interventional procedures. Funding will support regulatory approvals, clinical trials and commercialization across China, the US and Europe. [link]

AI-powered computer vision is transforming Las Vegas casino surveillance, detecting dealer errors, cheating and player behavior. Resorts World uses it on high-limit tables, while EagleSight.ai serves 12 properties. Accuracy reportedly improves from 40% initially to 90% within six months, with future potential for floor-wide monitoring and business intelligence. [link]

Kornia released Kornia-SLAM v0.1.0, an open-source Rust-based real-time visual-inertial SLAM system supporting monocular, stereo and VIO ORB-SLAM, loop closure, optimisation, multiple sensor inputs and Copper Robotics integration. Introduced at IROS 2026, the early-stage project invites contributions across computer vision, SLAM, robotics and Rust. [link]

Lexus of Sarasota has adopted LumaScanner to automate vehicle inspections, capturing tire wear, body damage and panel condition in real time when vehicles arrive. The deployment gives service teams greater visibility and customers more confidence, with installation supported by Brett H. [link]

SiMa.ai raised $150M Series C at a $1.45B valuation, bringing total funding to $500M. Its Physical AI platform combines Palette Neat software with purpose-built silicon, targeting humanoids, autonomous vehicles and drones, with 1,000 dense TOPS hardware planned by 2028. [link]

Indiana University researchers won first place in ECCV’s Kilometer Audio challenge with AVRA, an agentic AI system that understands hour-long videos using audio, visual and text information. The system autonomously searches video content to answer complex questions, with potential applications in search-and-rescue, investigations, scientific analysis, video search and AI-powered memory. [link]

Researchers demonstrated that computer vision can estimate gait in people with multiple sclerosis using a single consumer-grade video and MediaPipe Pose. In 153 participants, walking speed and stride time predicted T25FW performance, explaining 79% of variance. Longitudinal changes agreed 83% of the time, highlighting low-cost, accessible gait assessment without specialized equipment. [link]

At the Resource & Management Expo, Greyparrot CEO Mikela Druckman explains how computer vision enables continuous, real-time waste analysis in recycling facilities. Its data has gained Environment Agency acceptance, while Deep Nest helps brands redesign packaging for recyclability. The technology improves plant efficiency, supports EPR compliance, and shifts workers toward higher-value, data-driven roles. [link]

In a recent Ride AI podcast, Pony.ai CFO Leo Wang says Robotaxi operations are approaching sustainable economics, with nearly 2,000 vehicles, breakeven in Guangzhou and Shenzhen, and 691.2% YoY Q2 revenue growth. Its Gen-7 fleet is scaling while AI-driven development reduces costs. Internationally, Zagreb testing supports plans with Uber to deploy 2,000+ Robotaxis across Europe. [link]