Robotics

Today's Stories

1. 1

   AIxCrypto launches robot rental marketplace, aims to extend asset lifespan beyond sale

   Nasdaq-listed AIxCrypto unveiled RoboShare, a robot rental marketplace, and AIXC01, an infrastructure platform for autonomous assets, at the Automate Show in Chicago. RoboShare will launch publicly on June 22 in Los Angeles before expanding to New York and other cities, matching businesses seeking to rent robots with owners looking to monetize underused equipment. The robotics industry has historically generated revenue only at the point of sale. AIxCrypto's strategy aims to create ongoing value throughout a robot's operational lifetime through higher utilization rates, longer operating lives, and accumulated operational data—potentially lowering the cost of accessing robotic systems for businesses while creating new revenue streams for robot owners.

   The company plans to develop inspection, valuation and recirculation standards for pre-owned robots to support a secondary market in used robotic systems. It also intends to apply the same framework to autonomous aerial systems, including drones, creating common infrastructure for both ground-based and airborne autonomous assets.

2. 2

   Lumos Robotics' 2.8B model tops global embodied AI benchmark

   Lumos Robotics announced that its Prime R0 model—a 2.8-billion-parameter AI system—achieved the highest overall score on the MolmoSpaces leaderboard for zero-shot embodied AI, outperforming larger models from Nvidia and research teams at MIT and Princeton despite using less than one-sixth as many parameters. The result validates Lumos's strategy of designing AI for industrial deployment rather than building ever-larger foundation models. Prime R0 runs on consumer-grade hardware (Nvidia GeForce RTX 5060 8GB GPU), delivers millisecond-level inference times, and has been integrated into the Lumos Touch robotic arm for real-world tasks like floral arrangement, textile handling, and parts sorting—suggesting practical AI embodiment for manufacturing is becoming viable.

   Lumos plans to expand Prime R0 from manufacturing into logistics and additional sectors, positioning the underlying Lumos NexCore platform as a foundational operating system for next-generation industrial robotics.

3. 3

   Flytrex reaches 8,000 coordinated drone flights monthly in shared Dallas airspace

   Flytrex and Wing completed approximately 8,000 drone delivery flights in overlapping airspace between January and February 2026 across operational areas in Little Elm and Wylie, Texas, with zero reported airspace conflicts. The achievement represents scaling from a handful of overlapping flights to thousands per month in under a year, using the FAA's automated Unmanned Traffic Management (UTM) system to coordinate routes without manual intervention. This demonstration that competing drone operators can safely conduct simultaneous commercial delivery operations in the same airspace—using automated coordination rather than human oversight—may help establish the technical and operational foundation for multi-operator drone delivery to expand across U.S. metropolitan areas. The system successfully deconflicted every operational flight intent during the test period.

   Flytrex and Wing operated simultaneously on 30 of 31 active days during the test, with more than 10 hours of overlapping daily operations, and combined daily operations increased by 215 percent between January and February. The FAA's UTM Operational Evaluation currently includes 17 UTM service providers and drone operators, with the framework intended to explore how autonomous traffic management could support the growing number of low-altitude aircraft operating in shared airspace.

4. 4

   HelloFresh expands chilled capacity fivefold with robot automation

   Locus Robotics warehouse automation, adapted for cold storage, enabled HelloFresh's Factor brand to expand chilled fulfillment from 100 SKUs to 500 SKUs. Factor deployed 13 robots in July 2025, then added 26 more within three months, with plans to support EveryPlate fulfillment later this year. Temperature-controlled fulfillment is operationally demanding; Locus robots completed order cycles in 3 minutes and 36 seconds on average, compared with typical third-party logistics benchmarks of 30-minute, 60-minute, or multi-hour windows. This speed and capacity expansion enables HelloFresh to offer greater meal variety without the inflexibility and single-point-of-failure risk of fixed conveyor systems.

   Locus Robotics developed a heated motor enhancement and charging modifications to ensure reliable operation inside chilled environments, where battery efficiency typically degrades. The company now supports approximately 12,000 square feet of HelloFresh's chilled fulfillment space across two high-speed meal kit picking lines.

5. 5

   Aptiv's AI perception system chosen for Robust.AI's Gen 3 Carter warehouse robot

   Robust.AI has selected Aptiv's Pulse sensor—an AI-powered perception system combining radar and vision—for its Gen 3 Carter collaborative mobile robot. Aptiv is also working toward PL(d) safety certification for Pulse across relevant industrial safety use cases. Warehouse automation requires perception systems that work reliably in real-world conditions such as dust, glare, moisture, and reflective surfaces that degrade conventional systems. By combining Aptiv's sensor fusion with Robust.AI's AI perception and navigation technology, the partnership aims to deliver safer, more scalable warehouse automation while meeting functional-safety standards increasingly demanded by the market.

   The companies are advancing Pulse toward PL(d) certification—a high-reliability safety classification under ISO 13849-1 standard for hazardous robotics applications. This certification pathway reflects growing industry demand for functional-safety support as robots operate with higher automation near people and equipment.

6. 6

   RLWRLD launches open platform to benchmark robotic hands

   RLWRLD, a physical AI company, unveiled All Hands Up!, a free web platform that shares real-world performance data and design trade-offs of dexterous robotic hands. The platform includes data on more than 10 robotic hands and offers interactive visualization tools and side-by-side comparisons accessible directly in a web browser. Robotic hands are a core component of physical AI, but no single product today can satisfy all requirements due to inherent trade-offs between size, grip force, and back-drivability (the ability to respond compliantly to external forces). The platform addresses a question repeatedly raised in research and industry: which robotic hands perform effectively in real-world environments. This shared reference point may help manufacturers validate designs and help researchers and partners set clearer adoption criteria.

   RLWRLD proposes a dual-hardware strategy—Type 1 for field deployment (lightweight, durable) and Type 2 for training data collection (high back-drivability, precision)—reflecting that no perfect robotic hand yet exists. The company plans regular quarterly content updates to the platform.