Dreame brought its Echo service robots to this year’s IFA show in Berlin with a practical goal: using physical artificial intelligence to handle everyday chores. The technology combines the ability to perceive surroundings, interpret tasks, and act in the physical world.
Robots that dance, run, and cross obstacles demonstrate progress. Moving from exhibition stands into homes requires something more: completing useful work reliably.
Some physical AI products remain proof of concept or demonstrate only isolated capabilities. Fluid movements may be programmed routines, and prototypes assembled from open-source tools still need testing under continuous, complex workloads. As commercialization approaches, engineering reliability matters as much as the demonstration.
The Echo series explores that challenge through two designs: a single-arm robot focused on laundry and a two-armed model intended for broader services.
Completing a chore, from start to finish
Dreame’s robot vacuum business began its service robot development with a specific chore.
The Echo S1 is a wheeled, single-arm robot designed to handle laundry. Its tasks include collecting clothes scattered around the home, pulling laundry baskets, and interacting with washers and dryers. The aim is to coordinate the full routine, from gathering garments to moving them through washing and drying.
Completing that sequence requires more than mastering a grasping motion.
A home is not a laboratory with a limited set of variables. Clothes are soft, change shape, and do not stay in fixed positions. Lighting, obstructions, friction, and collisions can affect operations at any moment. The robot must recognize clothes and appliances, locate them in space, grasp flexible objects, interact with equipment, and coordinate each stage of the process. An error at any point can interrupt the entire task.
The S1 is expected to use two foundation models. A multimodal large language model (MLLM) will handle real-time voice interaction in multiple languages and dialects. A vision-language-action (VLA) model will combine spoken instructions with visual information to translate tasks into movements.
The Echo S1’s arm can be raised and lowered, and the robot’s overall height is adjustable from 1,570 to 2,275 millimeters. Its shoulder-to-wrist reach is 1,150 millimeters, and the arm can carry up to five kilograms. These specifications are intended to help it work with washers and dryers of different heights, collect clothes, pull baskets, and transfer garments. The single-arm design reflects a tradeoff among efficiency, reliability, and cost.
The two-armed Echo P1 explores a broader range of services, with the longer-term goal of an all-purpose robotic butler.
Its coordinated arms, omnidirectional mobile base, and modular design are intended for commercial tasks such as preparing made-to-order drinks, picking up and placing merchandise, and delivering on-demand retail orders. Potential household uses include organizing and moving objects.
What robot vacuums bring to service robots
Dreame’s robot vacuum business gives it experience to draw on as it develops service robots.
Robot vacuums were among the first widely used robotics products. In homes, they map spaces, plan routes, avoid obstacles, clean, and return to charge. Those functions depend on computer vision, navigation and motion control, precision mechanics, and energy management.
According to IDC data, Dreame’s robot vacuum business accounted for 26.1% of global unit sales and 31.2% of sales revenue in the first half of 2026, ranking first on both measures. That represents roughly one in four robot vacuums sold worldwide. The large installed base also provides feedback from household use that can help refine its technology.
The Echo S1 and P1 share capabilities in embodied intelligence models, robotic arm control, and computer vision. Dreame can also draw on its R&D, manufacturing, supply chain, and reliability testing systems to develop prototypes into finished products.
Experience in homes can help identify problems that are easy to miss in a laboratory: differences in room layouts, users’ needs, and faults that emerge only after prolonged use. That knowledge gives Dreame a basis for choosing which daily tasks its service robots should tackle.
Dreame is pursuing focused tasks and broader services in parallel. The S1 targets collecting and moving clothes and coordinating washing and drying. The P1 explores moving objects, tidying spaces, AI-assisted care, and security patrols. The two approaches extend its work into both household and commercial settings.
Household service robots are still at an early stage, however. Homes are complex and changeable environments, with demanding requirements for safety, stability, cost, and certification.
Dreame plans to build experience and improve capabilities gradually before moving into more complex household settings. That requires coordinating perception, decision-making, movement, manufacturing, and testing in real environments. Combining an algorithm, a model, and hardware is only part of the work.
Working toward practical deployment
The next challenge is sustained operation outside controlled demonstrations.
Complex chores must be broken into individual actions, those actions linked into complete tasks, and laboratory results tested through prolonged use in homes.
Dreame’s focus on laundry gives that work a defined starting point. It allows the company to identify problems in a familiar routine while developing capabilities for more varied services.
Moving from cleaning floors to handling clothes and other objects will require sustained investment and testing. The Echo robots broaden Dreame’s product ambitions, but their usefulness will depend on how consistently they complete those jobs.
KrASIA features translated and adapted content that was originally published by 36Kr. This article was written by Xiao Xi for 36Kr.