Warehouse automation is gaining momentum as robotics companies seek to overcome labor shortages and increase operational safety. In an effort to advance real-time autonomy, Robotec.ai collaborated with Liquid AI and Advanced Micro Devices Inc. (AMD) to showcase a robot capable of dynamically responding to warehouse conditions. The demonstration included live interactions at ROSCon 2025 in Singapore, where experts and industry professionals observed the operational efficiency of agentic AI in action. Observers noted the potential of integrating next-generation robotics platforms into everyday warehouse environments, with performance metrics and system reliability being closely monitored.
Developments preceding this demonstration often centered on pre-scripted robotics and basic automation routines, which limited response to unplanned situations. Earlier showcases prioritized standardized robot behavior over context-aware integration of AI and real-world scenarios. The latest exhibition marked a shift toward embedded AI that processes visual and natural language data on-device. These advances differ from previous technology announcements, as focus has moved from simulation-only environments to practical, data-driven adaptation using physical robots, marking a significant transition in application and scalability for broader industrial use.
How Does Agentic AI Guide Warehouse Robots?
Robotec.ai demonstrated a fully autonomous warehouse robot using AMD Ryzen AI processors and Liquid AI’s LFM2 foundation model. The robot interprets human commands, detects safety risks such as spills or blocked exits, and autonomously executes corrective measures within the warehouse. By processing data on-device, the system avoids reliance on traditional pre-programmed scripts. Such decentralized decision-making enables the robot to adapt to shifting warehouse conditions, as well as interact naturally with human operators.
What Role Do AMD and Liquid AI Play?
AMD provides the hardware backbone through its processors designed for AI tasks, while Liquid AI supplies vision language models that combine perception, reasoning, and language understanding. According to Robotec.ai, simulation-driven synthetic data is utilized to fine-tune these models for industrial scenarios and promote system robustness. The embedded intelligence on AMD platforms allows the robots to perform goal-driven actions without cloud dependency, resulting in improved speed and power efficiency.
“Our system operates fast, compact, and efficient, with exceptional performance in speed and power efficiency,”
Robotec.ai stated.
How Does Simulation Advance Autonomous Robotics?
The project emphasized the importance of simulation in validating robotics performance before live deployment. Robotec.ai’s collaboration with AMD led to a hardware-in-the-loop (HiL) simulation setup running a ROS 2 stack, unifying robot control logic across both simulated and physical environments. This approach facilitates rapid prototyping and seamless transition from test scenarios to real-world operations.
“HiL simulation enables us to replicate and refine real-world performance, accelerating research and development for OEMs,”
a company representative explained.
Robotec.ai highlighted a user interface that displays robot plans, missions, and reasoning steps in real time to human operators, further advancing transparency in decision-making. The ongoing evolution from demonstration to deployment will include transitioning from AMD Ryzen to embedded x86 solutions, signifying an intent to broaden the scope of application. This direct monitoring capability aims to ensure safety and allow rapid intervention if anomalies arise.
Adoption of agentic AI robots on the warehouse floor represents a significant shift from static automation toward adaptable, context-aware systems. Companies implementing such robotic solutions should focus on scalability, appropriate hardware selection, and rigorous simulation validation for industrial reliability. While the demonstration points to progress, further study into safety, workforce integration, and real-world durability remains critical for long-term adoption. For logistics and warehouse operators considering the next generation of robotics, attention to on-device AI processing, simulation-supported development, and collaborative system design may yield improved efficiency and adaptability under real-world constraints.
