Transforming the robotics industry: A sustainable communication protocol

 

Open-source robotics insights

• A high-routing efficient protocol enables direct communication between devices on the same network, moving data through the shortest path available while reducing network hops.
• Zenoh was engineered to tackle the energy inefficiencies inherent in traditional communication technologies.

---

The first non-data distribution service (DDS) robotic middleware (RMW) alternative has been adopted by the Robot Operating System 2 open-source project (ROS 2). The Open-Source Robotics Foundation (OSRF) introduced this alternative to address communication technical limitations, improve the out-of-the-box experience for new users and create a more robust RMW abstraction layer. However, little has been said about how this shift will impact the industry’s overall sustainability by addressing long-standing energy inefficiencies of established technologies.

Let’s explore how Zenoh, the communication protocol chosen as the new RMW, contributes to this energy transition and what this means for robotics.

Three key elements of energy efficiency of communication protocols

When examining the energy efficiency of communication protocols, three key elements are important. The first, known as processing efficiency, measures the amount of work required to create and process a packet during user-data transmission and reception. The second, referred to as wire efficiency, quantifies the additional bytes a protocol adds on top of user data. The third, routing efficiency, considers the distance –– or network hops –– data must travel to reach its destination.

Traditional protocols like the data distribution service (DDS), which have been the backbone of ROS 2 since its creation, come with significant wire overhead. DDS adds 56 bytes on top of every data message to ensure participant discovery, reliable delivery, message order, Quality of Service (QoS) policies, and more. While necessary for its proper functionality, excessive overhead slows down communications and reduces wire efficiency. Intuitively, moving large amounts of overhead bytes across the network leads to an increase in energy consumption, which makes operating a robotic system more costly.

Conversely, protocols like message queuing telemetry transport (MQTT) require all data exchanged by the applications to go through a cloud-based broker, which might be located far away from the devices, even for local communication. A high-routing efficient protocol would enable direct communication between devices on the same network, moving data through the shortest path available while reducing network hops. Lamentably, the topology constraint of MQTT results in the data having to be relayed and traveling longer distances across the network, enlarging energy expenditure.

The wide industry adoption of protocols like DDS and MQTT has made the communication stack of robots a power-hungry endeavor. For instance, many Autonomous Mobile Robots (AMRs) can use up to 50% of their operational energy without even moving, spending their battery only on data communication and processing. For a modern AMR, this would be the equivalent of running a high-wattage appliance, such as a clothes dryer (about 3kW), continuously for eight hours without drying any clothes. Reducing protocol overhead and improving data routing are crucial steps toward minimizing energy waste and enhancing the overall performance of robotic systems.

Introducing Zenoh: An energy-efficient protocol

Zenoh is a novel pub/sub/query protocol designed to unify data in motion, data at rest and computations across the cloud to micro-controller continuum. It offers a set of unified abstractions and supports different network topologies like peer-to-peer, routed, mesh and brokered communication. Zenoh runs efficiently on both server-grade hardware and constrained microcontrollers and networks (see Figure 1), featuring a minimal wire overhead of 5 bytes, low latency and high throughput performance.

Zenoh has been engineered from the ground up to tackle the energy inefficiencies inherent in traditional communication technologies. Unlike DDS, Zenoh reduces the energy required for discovery data transfer in robotics applications by up to 90% (see Figure 2). This reduction is achieved through a combination of minimal overhead and an innovative discovery approach. While DDS adds 56 bytes per message, Zenoh lowers this overhead to as little as 5 bytes. Additionally, while DDS’s dynamic discovery information provides very precise and extensive details about participants’ availability in the system –– quickly leading to scalability issues –– Zenoh employs sets and set-theory operations to generalize information distributed across the network, vastly reducing discovery traffic and enhancing wire efficiency.

In addition to this, unlike MQTT, one of Zenoh’s standout features is its ability to facilitate local communication by not imposing any constraint on the network topology of the system, allowing devices to communicate directly or through minimal intermediary steps without the need of a broker. This locality allows Zenoh to minimize network hops, achieving not only faster but also far less energy-intensive communication. Zenoh’s wire and routing efficiency enables it to perform exceptionally well on both constrained hardware and low-power networks.

What about the processing efficiency? According to a study by National Taiwan University (NTU) researchers, Zenoh’s throughput is largely higher than other protocols –– 10 times greater than MQTT’s and approximately five times higher than DDS’s. In a single-machine scenario, and similarly for a multi-machine one, Zenoh in peer-to-peer mode outperforms DDS by roughly 2x, MQTT by 130x, and Kafka by 65x at an 8-byte payload size. This means Zenoh requires significantly less central processing unit (CPU) power to handle the same amount of data.

Zenoh’s impact on robotics and ROS 2

Zenoh’s flexibility in supporting various communication topologies, combined with its ability to manage high throughput with low latency and minimal overhead, made it the preferred RMW alternative for ROS 2. But beyond this, Zenoh also addresses many other long-standing challenges in robotic system deployments, particularly those critical for robot-to-anything (R2X) applications. Imagine a fleet of robots operating in a warehouse managed by a Robotics Fleet Management (RFM) system. In this scenario, real-time communication between robots and the cloud is essential for efficiently managing tasks such as transporting goods. Traditionally, this would involve handling heterogeneous networks and protocols while ensuring scalability, a complex and cumbersome task for developers.

Zenoh facilitates easy integration by supporting a broad range of network technologies, including transmission control protocol/internet protocol (TCP/IP), user datagram protocol/IP, quick UDP internet connections (QUIC), Serial, Bluetooth and more. Its extensive plug-in ecosystem allows for interoperability with other existing protocols in the risk management framework (RMF) system without significant modifications. Furthermore, Zenoh’s capability to scale is particularly beneficial for deployments requiring robust communication across large robot fleets. With a unified application programming interface (API), it enables developers to focus on creating a more responsive RMF and fleet of robots more simply, rather than dealing with network complexities (see Figure 3).

Moreover, Zenoh’s proper use of resources aligns well with the growing emphasis on sustainability in technology. Reduced energy consumption translates directly into lower operational costs and a smaller carbon footprint, making R2X systems more environmentally friendly. For instance, in a comparison of communication costs over a satellite link between Zenoh and DDS, the former resulted in being approximately 92% cheaper over a year than the latter, a substantial cost save thanks to the protocol’s efficiency [X6]. Now that Zenoh is available as an RMW, all ROS 2-based systems can immediately tap into its full range of features and leverage its clear energy advantages.

The future of robotics with Zenoh

The shift to Zenoh in the open-source robotics community not only enhances the viability of the systems but also paves the way for innovations. By addressing the critical challenges of traditional communication protocols, developers can focus on creating more advanced and capable robots without being hindered by the limitations of the existing technologies. Its adoption marks a significant milestone in driving the robotics space towards a more efficient, cost-effective and feasible future, not only benefiting the industry but also contributing to broader efforts to promote environmental sustainability in technology.

Do you have experience and expertise with the topics mentioned in this content? You should consider contributing to our WTWH Media editorial team and getting the recognition you and your company deserve. Click here to start this process.