Robot Communication Protocols: Why EtherCAT and CAN Are the Real Future？

As the robotics industry continues to evolve at high speed, communication protocols are no longer just about “connecting devices.” They directly determine system performance, architectural complexity, scalability, and long-term competitiveness.

Although many industrial communication protocols exist today, a clear consensus is emerging within the robotics community: EtherCAT and CAN are becoming the two most critical and indispensable protocols for future robotic systems.

1. EtherCAT: The De Facto Standard for High-Performance Robot Control

1.1 Growing Market Dominance

According to statistics released by the EtherCAT Technology Group (ETG) in 2024, EtherCAT has captured approximately 39.2% of the global industrial robot communication protocol market, with an annual growth rate of 12.7%, significantly outperforming other industrial protocols.

EtherCAT has become the default choice in many mission-critical scenarios, including:

• Multi-axis synchronous control in industrial robots
• Real-time coordination of humanoid robot joints
• Autonomous driving and multi-sensor fusion
• Human–machine collaboration in Industry 4.0 environments

Leading robot manufacturers such as KUKA and FANUC widely adopt EtherCAT as their primary control bus for complex operations such as welding, material handling, and painting.

1.2 Designed for Real-Time Motion Control

Robot joint control typically relies on current, velocity, and position loops, forming a closed-loop control system with extremely demanding communication requirements:

• Ultra-low latency
• Strong determinism
• Precise multi-axis synchronization

EtherCAT was designed specifically to meet these demands.

Key technical advantages include:

• Distributed Clock (DC) technology
• Highly streamlined protocol stack
• Exceptional bandwidth utilization

1.3 “Processing on the Fly”: EtherCAT’s Core Moat

One of EtherCAT’s most distinctive innovations is On-the-Fly Processing.

Unlike traditional Ethernet protocols that rely on store-and-forward mechanisms:

• EtherCAT slaves process data while the frame is passing through
• No need to buffer the entire frame
• Per-node processing delay is typically around 1 μs

This design enables EtherCAT to maintain deterministic real-time performance even in large, multi-axis systems. At present, this mechanism is essentially unique to EtherCAT.

1.4 Why EtherCAT Became the Mainstream Choice

From a pure performance and safety perspective, EtherCAT is exceptionally strong. However, its widespread adoption is driven not only by performance, but also by its openness and long-term cost efficiency.

While EtherCAT may present a higher learning curve for engineers compared to CAN, it is almost irreplaceable in scenarios involving:

• High-performance motion control
• Tight multi-axis synchronization
• Large-scale distributed robotic systems

In applications where motion performance matters, EtherCAT often delivers the best overall cost-to-performance ratio over the system’s lifecycle.

2. CAN: The Unavoidable Foundation of Robotic Communication

2.1 Still Indispensable in Real-World Systems

CAN (and its motion-oriented variants such as CANopen and CAN FD) remains another dominant communication solution in robotics, particularly in scenarios where real-time constraints are less stringent:

• Lower-body control of humanoid robots
• Wheeled robots and mobile platforms
• Quadruped robots, AGVs, and service robots

Even as EtherCAT hardware costs continue to decline, CAN remains deeply embedded in many robotic architectures.

A common system design in humanoid robots is:

• Upper body: EtherCAT (high-speed, synchronized motion)
• Lower body or subsystems: CAN / CAN FD

2.2 Core Strengths: Reliability, Simplicity, Robustness

Originally developed for automotive electronics, CAN was designed with a strong emphasis on:

• High reliability
• Excellent noise immunity
• Distributed decision-making

CAN uses a CSMA/CA non-destructive arbitration mechanism:

• Multiple nodes may attempt to transmit simultaneously
• Higher-priority messages continue uninterrupted
• Lower-priority messages automatically defer without data loss

This makes CAN especially suitable for transmitting:

• Digital I/O signals
• Sensor data
• Non-strictly periodic control information

2.3 Limitations and Practical Boundaries

As robotic systems scale up, CAN’s inherent limitations become more visible, especially when:

• The number of joints exceeds dozens
• Control frequencies increase
• Strong multi-axis synchronization is required

Although CAN FD and multi-segment CAN architectures can mitigate some issues, EtherCAT remains superior for high-performance, tightly synchronized motion control.

In practical engineering terms, the rule of thumb is clear:

• Few axes, moderate performance, existing CAN infrastructure → CAN is sufficient and cost-effective
• Many axes, high dynamics, system-level synchronization → EtherCAT is the better choice

3. I3C: A Promising but Emerging Contender

I3C is a relatively new sensor communication protocol that is gaining attention in robotics, particularly in dexterous hands and high-density sensor applications.

3.1 Potential Advantages of I3C

• No external PHY required, simplifying hardware design
• Supports multiple devices on a single bus
• Well suited for space-constrained, high-density sensor layouts

Major semiconductor vendors are actively promoting I3C support, including:

• NXP (i.MX RT1180)
• Infineon (PSOC Edge)
• Renesas (RA8 series)
• Microchip (PIC18-Q20 series)
• STMicroelectronics (STM32H / STM32N / STM32U families)

3.2 Reality Check: Not Yet at Scale

Despite its promise, in real robotic deployments—especially dexterous hands:

• CAN FD remains the dominant solution
• I3C ecosystems are still immature
• Concerns remain regarding noise immunity in harsh industrial environments

That said, as domestic and international chip development continues, and with emerging protocols such as CAN XL, the communication landscape may evolve further in the coming years.

Conclusion: Not Replacement, but Clear Role Separation

From a practical engineering standpoint, the future of robotic communication will not be defined by a single protocol replacing all others. Instead, it will be characterized by clear division of roles:

• EtherCAT as the backbone for high-performance, synchronized motion control
• CAN as a robust, cost-effective foundation for distributed communication
• I3C gradually exploring niche applications in sensors and compact actuators

Ultimately, protocol selection depends on performance requirements, system scale, cost structure, and long-term evolution strategy.

What is already certain is this: 👉 EtherCAT and CAN have become the most stable and reliable foundations of modern robotic systems.

The ARMxy series ARM embedded computers support real-time Linux and EtherCAT communication via the IGH EtherCAT master stack, delivering response times of around 100 μs. With built-in CAN interfaces, ARMxy provides native support for CAN-based robotic and industrial control systems.