A New Direction for Industrial Automation
Vention is set to showcase a broader AI-driven automation strategy at IMTS 2026 by combining Physical AI and Agentic AI on one industrial automation platform. The company will demonstrate how AI can support both machine-level operations and system-level engineering workflows.
This approach moves beyond using AI as a standalone software tool. Instead, Vention positions AI across the automation lifecycle, from cell design and programming to robotic motion, deployment, monitoring, and troubleshooting.
Two Layers of Intelligence for Modern Manufacturing
Vention's AI-Defined Automation strategy combines two different but connected forms of intelligence.
Physical AI operates directly at the machine level. It enables robots and automated equipment to perceive objects, interpret their environment, plan movements, and respond to changing production conditions.
Agentic AI operates at the system and engineering level. It uses natural-language interaction and AI agents to assist engineers with automation design, industrial programming, machine analysis, and operational troubleshooting.
In my view, this combination addresses an important limitation in current industrial AI adoption. Many AI solutions improve either engineering workflows or machine performance. Connecting both layers within the same automation architecture could reduce the gap between digital engineering decisions and physical machine behavior.
MachineMotion AI Provides the Computing Foundation
At the center of Vention's IMTS demonstrations is MachineMotion AI, the company's next-generation automation controller.
The controller supports robotics, motion control, AI vision, and automation applications while connecting software and physical equipment through a unified architecture. Vention states that the platform uses NVIDIA Jetson, NVIDIA Isaac, CUDA-accelerated libraries, and open AI models such as NVIDIA FoundationPose.
This architecture reflects an important industry trend. Industrial controllers are increasingly expected to process more than deterministic control logic. They must also handle vision data, AI models, motion planning, and communication with cloud-based engineering tools.
MachineAgent Expands AI into Engineering Workflows
One of the main IMTS 2026 demonstrations will feature MachineAgent, Vention's Agentic AI technology.
MachineAgent uses plain-language prompts to support automation cell design, industrial program generation, machine deployment, and operational analysis. The system is designed to help users move from an engineering concept toward a working automation application through a more connected workflow.
Visitors will also see MachineLogic Copilot used to program a physical UR3 robotic cell through Vention's cloud platform. In addition, the company plans to demonstrate an AI-assisted developer workflow using Claude Code, Vention CLI, and Developer Toolkit 2.0.
The engineering value of this approach depends on how effectively AI-generated outputs can integrate with established industrial validation processes. AI can accelerate engineering work, but manufacturers will still need simulation, testing, safety verification, and controlled commissioning before deployment.
Physical AI Targets Robotic Adaptability
Physical AI will also play a major role across Vention's IMTS demonstrations.
Traditional robotic systems often require engineers to define precise positions, motion paths, and handling conditions in advance. However, real production environments frequently introduce variation in part location, orientation, and surrounding equipment.
Vention's Physical AI technology aims to address these conditions through AI vision, object recognition, pose estimation, intelligent motion planning, and autonomous decision-making.
This could be particularly useful in applications where fixed programming becomes difficult to maintain, including bin picking, machine tending, assembly, and material handling.
Rapid Operator AI Demonstrates Deep Bin Picking
Vention will demonstrate Rapid Operator AI with a UR12e robot performing deep bin picking.
The application combines Vention's GRIIP software with NVIDIA Isaac foundation models to support real-time part detection, 6-DoF pose estimation, collision-free motion planning, and adaptive retry logic.
According to Vention, the demonstration can achieve up to a 99% first-pick success rate under its demonstrated conditions.
Deep bin picking remains one of the more demanding robotic applications because robots must deal with overlapping parts, changing orientations, occlusions, and uncertain grasping conditions. Therefore, improvements in perception and motion planning could help reduce the engineering effort required for these systems.
Autonomous Collision-Free Path Planning
Another demonstration will feature a FANUC LR Mate robot using AI-driven collision-free path planning.
The system combines on-arm vision, a digital twin, and AI motion planning to generate robot paths without requiring engineers to manually program every intermediate waypoint.
This development is significant because robot programming often becomes time-consuming when equipment layouts change. A system that can identify a target and calculate an acceptable motion path could shorten adjustment and commissioning time.
However, autonomous path generation must still operate within defined safety limits and production constraints. For industrial deployment, AI-generated motion must remain predictable, validated, and compatible with the surrounding safety architecture.
MachineAgent Connects Design, Programming, and Operation
Vention will demonstrate MachineAgent with a UR3 robotic cell across three stages: Design, Program, and Operate.
The workflow includes AI-generated automation cell layouts, industrial programming, digital twin simulation, physical deployment, fleet analysis, and troubleshooting.
This end-to-end structure may become increasingly important as manufacturers manage larger numbers of automated systems. Instead of separating mechanical design, controls engineering, simulation, deployment, and maintenance into disconnected tools, unified platforms can create a more continuous engineering workflow.
The real advantage may not come from AI replacing automation engineers. Instead, AI could allow engineers to spend less time on repetitive configuration and data analysis while focusing more on process optimization and system validation.
Additional Robotics and Motion Demonstrations
Vention will present several additional automation applications at IMTS 2026.
The Overhead Range Extender with a UR20 robot will demonstrate a seventh-axis system in a live welding application. The configuration is designed to extend robot reach for welding, machine tending, assembly, and long conveyor applications.
The MachineMotion AI Daisy Chain demonstration will show how one controller can support multiple motors and actuators through a cabinet-free architecture. Vention states that the system can daisy chain up to 20 motors.
Another demonstration will feature Click & Customize Machine Tending with a FANUC CRX-10iA robot. The application focuses on configurable CNC loading and unloading programmed through MachineLogic.
Together, these demonstrations show that Vention is applying AI across different levels of industrial automation rather than limiting it to robotic vision alone.
Why the Physical AI and Agentic AI Combination Matters
The industrial automation sector is moving toward more software-defined and data-connected architectures. However, increased software capability also creates greater engineering complexity.
Physical AI can help machines respond to changing physical conditions. At the same time, Agentic AI can help engineers and operators manage the increasing amount of information generated across automation systems.
In my opinion, the strongest part of Vention's strategy is the attempt to connect these two areas. Machine intelligence becomes more useful when engineering tools understand operational data, while engineering AI becomes more valuable when it can work with the real behavior of physical equipment.
The long-term challenge will be maintaining deterministic control, functional safety, cybersecurity, and engineering accountability as AI takes a larger role in industrial automation.
Vention Prepares for IMTS 2026
Vention will exhibit at IMTS 2026 from September 14 to September 19 at Booth 236860 in the North Hall.
The company will also host a presentation titled From Code to Concrete: How Cloud Robotics and Physical AI Are Rewriting Industrial Automation. The session will explore how cloud-based robotics and Physical AI are changing the way manufacturers design and deploy automation systems.
IMTS 2026 will provide an opportunity to evaluate how far AI-driven industrial automation has progressed from concept demonstrations toward practical engineering applications.
Conclusion
Vention's IMTS 2026 presentation highlights a growing shift in industrial automation. AI is moving beyond isolated analytics and vision systems and entering the complete automation lifecycle.
By combining Physical AI for robots and machines with Agentic AI for engineering and operations, Vention is developing a more connected automation environment.
The technology still faces important industrial requirements, including validation, safety, deterministic performance, and maintainability. Nevertheless, the direction is clear: future automation platforms will increasingly combine control systems, robotics, AI perception, engineering software, simulation, and operational intelligence within the same ecosystem.