{"ok":true,"nota":{"id":"33f98722c2911516","slug":"mover-parte-del-cerebro-de-robots-a-la-nube-mejora-su-desempeno-en-tar-33f987","titulo":"Mover parte del cerebro de robots a la nube mejora su desempeño en tareas reales","resumen":"Microsoft Research encontró que trasladar la inferencia de IA fuera del robot mejora la tasa de éxito en tareas físicas, la eficiencia y la capacidad de ejecutar modelos más avanzados. El estudio evalúa robots en condiciones reales.","fuente":"microsoft-research","fuenteNombre":"Microsoft Research","color":"#00a4ef","autor":"Ganesh Ananthanarayanan, Matthew Balkwill, Xenofon Foukas, Sanjeev Mehrotra, Bozidar Radunovic, Connor Settle, Ankit Ver","url":"https://www.microsoft.com/en-us/research/blog/offloaded-inference-for-real-world-physical-ai-robotics/","imagen":"https://www.microsoft.com/en-us/research/wp-content/uploads/2026/09/SO101-handover-1.mp4","tema":"tecnologia","subtemas":["robótica","inteligencia artificial","computación en la nube","eficiencia energética","robots físicos"],"entidades":["Microsoft Research"],"formato":"dato","entretenimiento":false,"pubTs":"2026-09-23T16:01:36.000Z","indexTs":"2026-10-22T10:38:41.125Z","hash":"12a2ffb64e8d59f0c370dd7ae216d7b2763055298436eb9e892e9f367f8cb656","tsa":{"time":"Oct 22 10:39:39 2026 GMT","provider":"freetsa.org","sobre":"sello"},"stats":{"views":0,"likes":0,"clicks":0,"dwell":0},"tipo":"normal","registro":"hallazgo","porQueImporta":"Un robot con todo el procesamiento a bordo está limitado por la batería, el calor y el tamaño del hardware. Si parte de ese cálculo se hace en servidores remotos, el robot puede ser más liviano y hacer tareas más complejas. El costo es depender de una conexión estable, algo que no siempre existe en fábricas, campos o zonas con mala señal. Esa dependencia es el punto que habrá que resolver antes de llevarlo a lugares como los nuestros.","quienFirma":"Microsoft Research","confianza":"preprint","consulta":null},"sello":{"hash":"12a2ffb64e8d59f0c370dd7ae216d7b2763055298436eb9e892e9f367f8cb656","sello":"68bf65eb7bd0ac202f3ca9f2bb1b33d3adf49c1f2b592497d6d38eda4086211c","beacon":{"drand":{"cadena":"8990e7a9aaed2ffed73dbd7092123d6f289930540d7651336225dc172e51b2ce","ronda":6574484,"aleatorio":"17f8dc6691055e17d3340945fc697b1d14385d774dc400926b289d0664c160fb","tiempo":"2026-10-22T10:39:00.000Z"},"nist":{"cadena":2,"pulso":1994441,"salida":"917B00588BBAF41DD8C461E7D93EA47DC73C00983776285722380061C2505D68D1C49D178A1E9E1A422EEA35DB6367141AB79828DD0F9592CE1805F752B02469","tiempo":"2026-10-22T10:32:00.000Z"},"noAntes":"2026-10-22T10:39:00.000Z"},"verificacion":{"drand":true,"nist":true},"tsa":{"time":"Oct 22 10:39:39 2026 GMT","provider":"freetsa.org","sobre":"sello"},"original":{"titulo":"Offloaded inference for real-world physical AI robotics","descripcion":"Robots are getting smarter, but how can their hardware match that growth? New Microsoft Research findings show that moving AI inference beyond the robot can improve task success, boost efficiency, and support more advanced physical AI workloads. The post Offloaded inference for real-world physical AI robotics appeared first on Microsoft Research .","pubDate":"Wed, 23 Sep 2026 16:01:36 +0000","url":"https://www.microsoft.com/en-us/research/blog/offloaded-inference-for-real-world-physical-ai-robotics/"},"indexado":"2026-10-22T10:38:41.125Z"}}