Gasgoo Munich- PsiBot officially unveiled its in-house developed ψ-SynRobot on April 20, kicking off mass production immediately.
The ψ-SynRobot is designed as a general-purpose robotic platform for complex logistics, retail, and industrial environments. Unlike traditional industrial robots, which are often confined to specialized, closed systems, this model features an innovative architecture that integrates task execution with data collection. It not only handles high-precision operations with accuracy but also gathers high-quality operational data while on the job in real-world settings. This creates a continuous evolution loop: application deployment, data accumulation, model iteration, and capability upgrades.
Image source: PsiBot
To achieve this, the ψ-SynRobot employs a multi-joint biomimetic structure for dexterous manipulation, boasting the power to execute long-horizon tasks. It handles multi-step operations within complex environments with an industry-leading accuracy rate. Additionally, its advanced spatial perception system allows for rapid adaptation across various scenarios—including warehouse sorting, retail guidance, and industrial assembly—truly delivering "one machine, multiple uses" to meet the flexible demands of different sectors.
Furthermore, PsiBot embedded data collection capabilities directly into the core architecture during the initial design phase of the ψ-SynRobot. Every time the robot completes a task, it simultaneously generates a packet of high-quality training data covering vision, tactile feedback, and motion trajectories.
The ψ-SynRobot also incorporates comprehensive safety design, enabling stable 24/7 continuous operation. This significantly boosts operational efficiency in real-world deployments like warehouses, supermarkets, and industrial production lines while lowering labor costs.
PsiBot noted that with the official release and mass production of the ψ-SynRobot, the company has established a complete product matrix ranging from an "embodied AI brain" to a "holistic execution platform." This portfolio covers various product forms, including wheeled robots and exoskeleton gloves.
Founded in 2024, PsiBot has focused on overcoming the challenges of dexterous manipulation in embodied AI from the outset. The company builds its software toolchain around a core end-to-end VLA model. By utilizing its proprietary Psi-SynEngine embodied data collection engine, it efficiently acquires real-world interaction data, significantly reducing collection costs.
Prior to this launch, PsiBot successfully completed small-scale scenario validation in actual client warehouses, achieving a substantive breakthrough in sorting efficiency. Building on this foundation, the company plans to build and establish China's largest dexterous hand dataset by 2026. This will lay a solid data foundation for embodied AI to transition from technological leadership to a commercial closed loop.
