Representative End User Clients
Representative Automation Clients
Representative Software Clients
So far on our journey exploring the new era of industrial robotics, Patrick has established the critical hardware foundation at the robotic edge, and I’ve connected it to the software and data intelligence layer powered by the Industrial Data Fabric. This exploration has made one thing abundantly clear: with both the physical "body" and the contextual "brain" in place, the old ways of classifying robots are no longer sufficient for strategic analysis in the age of AI.
A frequent and important question we get is, "Where do cobots (collaborative robots) fit into this?" The answer is crucial: "cobot" is not a physical form factor, but a critical set of capabilities enabled by our taxonomy's layers. A robot becomes a cobot not because of its shape, but because of advanced safety systems in its hardware (Part 2 of our taxonomy below) and intelligent perception in its software (Part 3 of our taxonomy below). This allows it to work safely alongside people, making them more productive by assisting with repetitive or physically demanding tasks, which can reduce injuries from lifting or repetitive strain. Our taxonomy is designed to analyze these enabling technologies that turn a standard robot form into an intelligent and collaborative partner.
To analyze this new landscape effectively, we need a new language, a new framework that respects the physical reality while embracing the software-defined future. That is why ARC is developing a Unified Taxonomy for Industrial Robotics and Physical Intelligence.
As part of our voyage of discovery, we have already updated our working taxonomy beyond the traditional IFR standards to be more inclusive of the form factors we see gaining prominence with our industrial customers. This is a living framework, and as we continue our research, we welcome feedback and dialogue on the categories and definitions we are building. Our goal is to create a shared language for the entire industry.
Our taxonomy is structured in three essential parts.
Part 1: The Robot Platform (The Physical Form Factor)
This is our starting point. The physical form of a robot still dictates its fundamental capabilities. A key distinction here is between the robot's form and its function. Part 1 defines the form. A robot's function (e.g., "security robot" or "cobot") is determined by the combination of its form with the hardware and software in Parts 2 and 3. While this list covers the most prevalent industrial forms, we also acknowledge other, less common types like spherical and cylindrical robots.
This brings us to a critical point in the current market debate. There is an intense focus on the Humanoid Robot, driven by its incredible technical challenge and futuristic appeal. However, for most industrial use cases, true innovation will come from a 'function over form' approach. We must avoid the trap of 'paving the cowpaths'—simply replacing a human with a human-shaped robot—and instead fundamentally rethink industrial processes to leverage the optimal robotic form, be it a drone, an AMR, or a multi-limbed articulated arm. Our taxonomy is designed to encourage this thinking by treating the Humanoid as one important form factor among many, not the inevitable endpoint for all tasks.
Articulated Robot:
Description: This is the most common type of industrial robot, resembling a human arm. It features a series of rotary joints (typically 4 to 6 axes, but sometimes more) that provide exceptional flexibility and a large, spherical work envelope. It is the form factor most frequently adapted for collaborative use.
Key Characteristics: High dexterity, long reach, suitable for complex paths and orientations.
Primary Applications: Welding, painting, assembly, machine tending, packaging, and palletizing.
AI Evolution: AI-powered vision is making these robots easier to program for complex assembly and inspection tasks that previously required extensive jigs and fixtures.
SCARA Robot (Selective Compliance Assembly Robot Arm):
Description: SCARA robots are designed for speed and precision in a single plane. They typically have four axes and are compliant in the X-Y plane but rigid in the Z-axis, making them ideal for vertical insertion tasks.
Key Characteristics: High speed, excellent repeatability, compact footprint.
Primary Applications: High-speed pick-and-place, assembly, packaging, and sorting, especially in the electronics and consumer goods industries.
AI Evolution: AI is enhancing their ability to handle product variations on the fly, moving from high-volume, low-mix tasks to high-mix scenarios without mechanical changeover.
Cartesian/Gantry Robot:
Description: These robots operate along three linear axes (X, Y, Z) using a rigid, overhead structure. They are defined by a rectangular work envelope and are highly scalable.
Key Characteristics: High precision over large areas, heavy payload capacity, easily customized.
Primary Applications: Palletizing, material handling for large or heavy objects, CNC machine tending, automated storage and retrieval systems (AS/RS).
AI Evolution: AI-powered fleet management is optimizing the path planning for multiple gantry robots operating in the same workspace, maximizing throughput and avoiding collisions.
Parallel Robot (e.g., Delta Robot):
Description: These spider-like robots use multiple arms connected to a single platform. This design makes them exceptionally fast and precise for light-payload applications within a dome-shaped work envelope.
Key Characteristics: Very high speed and acceleration, high precision.
Primary Applications: High-speed pick-and-place, packaging, and assembly, common in the food, pharmaceutical, and electronics industries.
AI Evolution: AI-powered vision guidance is essential for these robots, allowing them to pick items from fast-moving, unstructured conveyor belts.
Mobile Robots:
Description: This broad and rapidly evolving category covers all ground-based robots capable of navigation. It includes Wheeled Robots like traditional AGVs (Automated Guided Vehicles) that follow fixed paths and AMRs (Autonomous Mobile Robots) that navigate dynamically, as well as Legged Robots like quadrupeds that offer unparalleled mobility in unstructured environments and over difficult terrain.
Key Characteristics: Mobility, flexibility, scalability for logistics and inspection.
Primary Applications: Intralogistics, line-side replenishment, goods-to-person picking, remote inspection, security patrols.
AI Evolution: This category is almost entirely driven by AI. Advanced perception, navigation (SLAM), and fleet management AI are the core differentiators, enabling these robots to operate safely and efficiently in dynamic, human-populated environments.
Drones/Unmanned Aerial Vehicles (UAVs):
Description: These are aerial robots, typically multi-rotor helicopters, capable of three-dimensional movement. They bring robotic capabilities to applications where ground-based robots cannot go, offering a unique perspective for data collection and inspection.
Key Characteristics: Flight capability, remote sensing, access to elevated or dangerous locations.
Primary Applications: Industrial asset inspection (power lines, wind turbines, bridges), construction site surveying and progress monitoring, precision agriculture, and in-warehouse inventory management in high-bay facilities.
AI Evolution: AI is the core of an autonomous drone. It powers the autonomous navigation, collision avoidance, and, most importantly, the analysis of the data it collects. AI-powered computer vision can automatically detect faults, count inventory, or measure stockpiles from the drone's sensor feeds.
Humanoid Robot:
Description: While technically a form of mobile robot, we classify humanoids as a distinct category due to their unique strategic importance, technological complexity, and application focus. These robots are designed with a human-like form factor (two legs, two arms, torso, head) to operate in environments built for people, using human tools. They represent a vital collaboration tool as a cobot, helping workers to be more productive and eliminate or minimize the risk of injury by performing functions in worker-hostile environments, and repetitive and heavy tasks that can cause injuries, such as to the back and wrist.
Key Characteristics: Unique versatility, ability to navigate stairs and cluttered environments, potential to use human tools.
Primary Applications: Currently in pilot stages for logistics, warehouse tasks, inspection, and as a productivity and safety tool helping workers in manufacturing and hazardous environments.
AI Evolution: Humanoids are a primary platform for leveraging Physical Intelligence, entirely dependent on advanced AI for balance, bipedal locomotion, manipulation, and human-robot interaction.
Part 2: The Hardware Component Stack (The Physical Nervous System)
This is Patrick's domain, deconstructing the robot into its constituent technological parts. Think of this as the robot's physical "nervous system" and "skeleton"—the core components that enable perception, action, and computation. Understanding this layer is critical for assessing the performance, cost, and supplier ecosystem for any robotics solution. For cobots, this layer is especially critical, as it includes the advanced sensors that ensure human safety.
Controller: This is the robot's local brain. It’s the ruggedized, industrial-grade controller or edge compute platform, increasingly GPU-accelerated, that runs the real-time software for perception and motion control directly on the machine.
Sensors & Perception Hardware: These are the eyes and ears of the robot. This rapidly innovating space includes 3D vision cameras, high-resolution LiDAR, and force/torque sensors that provide the raw data needed to understand the physical world.
Actuators: These are the muscles of the robot. The motors, gears, servos, and drives that convert electrical energy into precise physical motion.
End Effectors: These are the hands of the robot. This category includes the grippers, welding tools, paint guns, and other specialized tools that allow the robot to physically interact with its environment and perform its task.
Mechanical Structure: This is the robot's skeleton. The physical frame, joints, and components that make up the robot's chassis or arm and define its reach, payload, and physical constraints.
Part 3: The Software & Intelligence Stack (The Embodied Mind)
This is where my coverage of Industrial AI software connects directly to the physical machine. It represents the robot's "embodied mind"—a collection of interconnected software capabilities where value is added cumulatively, culminating in true autonomy. The "collaborative" nature of a robot is largely defined here, in the safety-certified software and the AI models that can predict human intent and react safely.
Foundational Middleware (OS): This is the essential software plumbing. Dominated by the Robot Operating System (ROS/ROS 2), it acts as a vendor-neutral integration bus.
Center of Gravity: This capability is purely Edge. It runs entirely on the robot's onboard controller to manage real-time communication between hardware and software.
Core Robotic Capabilities: These are the fundamental "verbs" of robotics. This includes the software algorithms for Perception (understanding sensor data), Navigation (moving through space), and Manipulation (interacting with objects).
Center of Gravity: This is also firmly at the Edge. These perception and motion planning tasks are latency-sensitive and must run on the robot for safe, real-time operation.
AI & Simulation: This is where higher-order brain functions reside. It includes the AI models for complex tasks, reinforcement learning, and the "sim-to-real" training environments that teach the robot how to perform its tasks intelligently.
Center of Gravity: This capability is Hybrid. The computationally intensive AI model training and large-scale simulation happen in the Cloud, but the resulting, optimized model (inference) is deployed to run on the Edge.
Fleet Management & Orchestration: This is the enterprise command and control capability. This software manages, monitors, and coordinates entire fleets of robots, integrating them with higher-level systems like a WMS or MES.
Center of Gravity: This is Cloud-centric, but Distributed. The central management dashboard, analytics, and integration with enterprise systems are in the Cloud. However, there are often Edge components for local coordination between robots in a specific workcell to ensure resilience if cloud connectivity is lost.
Human-Robot Interface (HRI): This is the communication bridge to people. It includes everything from traditional teach pendants to modern low-code programming interfaces and, increasingly, natural language and Generative AI-powered "copilots."
Center of Gravity: This is fully Distributed. An operator might use a ruggedized tablet on the factory floor (Edge), an engineer might configure a robot from a web browser (Cloud), and a maintenance technician could use AR glasses that overlay data onto the physical robot (Hybrid).
Unified Taxonomy for Industrial Robotics and Physical Intelligence
This unified taxonomy gives us a powerful, structured way to analyze the market. With this framework as our guide, our voyage of discovery can now chart a course toward the key strategic questions facing the industry. This taxonomy represents our current understanding in a fast-evolving market. It is a living framework that we will adapt based on feedback from our customers and our ongoing research into the evolution of this technology and the market landscape.