Le 30/01/2023
ZettaScale Simplifies and Streamlines Communication for Autonomous Ag Robots
The technology innovator’s Cyclone DDS and Zenoh ensure faster, safer and more efficient connections between advanced agricultural machines, the data they collect and the people who operate them.
Sponsored - This publication contains sponsored content.
Agricultural robotics equipment is getting smarter by the day. A far cry from the old-school tractors of previous generations, the latest advanced technologies move autonomously, sense their surroundings, identify crops, weeds and pest problems, navigate complex terrain and collect data at an astounding rate. Some machines have even begun working in tandem, offering farmers a swarm of robots that can complete a wide range of tasks more efficiently than ever before.
All of these amazing technological feats have something in common: They require simplified, streamlined communication between machines, microcontrollers, data systems, human operators and more. ZettaScale, a leader in communication software for robotics and automotive applications, has developed solutions to meet the ag industry’s newfound and ever-changing needs. Cyclone, a data distribution service (DDS) implementation, and Zenoh, an advanced protocol for robot-to-anything (R2X) communication, are two solutions the company has implemented to help machines and their operators communicate more effectively.
“Think about robot communication as the nervous system of the human body,” says Angelo Corsaro, ZettaScale’s Chief Executive Officer and Chief Technology Officer. “Nobody notices it when everything is working. When it isn’t, it’s a complete mess because nothing works. ZettaScale is the neuroscientist. We build robust and very efficient nervous systems.”
A person with a malfunctioning nervous system might have vision issues, trouble moving properly or an increased likelihood of burning their hand on the stove. The same could be said for agricultural machines. Instead of operating easily within their environments, the robots might misidentify crops, slam into one another or travel outside the boundaries of where its safe to operate. Cyclone DDS and Zenoh were developed to overcome these common communication issues and mitigate potential problems down the line.
Cyclone DDS acts as a default application for secure communications within agricultural robots.
The first thing to understand about Cyclone is that the DDS standard has been around for more than two decades. Many of the products that implemented this standard have been around much longer. Cyclone DDS is the culmination of twenty years of observing changing conditions, deepening the understanding of what users need and building a superior solution that meets those needs.
The current product was developed from the ground up. It now offers users a highly secure, open-source OMG DDS implementation that has been proven to perform in even the toughest real-world scenarios. Cyclone DDS is also certified to run in the most stringent safety critical environments. ZettaScale is proactively working toward achieving the ISO-26262 certification level for both Cyclone and Zenoh. The company believes this certification will be required sooner rather than later.
ZettaScale has noticed an important regulatory shift that indicates agricultural robots are beginning to require certifications similar to those within the automotive industry. It’s one thing to use an autonomous ag robot in a controlled environment on a single plot of land. For farmers who want to send their machines across the road from one field to another, these safety certifications will be necessary.
Zenoh provides efficient, scalable communication between robots, data centers, microcontrollers and more.
If Cyclone DDS is ZettaScale’s solution for in-robot communications, Zenoh is the innovative protocol that handles the way robots talk to everything and everyone else. According to Corsaro, there were a few major pain points that led to Zenoh. One occurred when users began to introduce collaborative machines or robot swarms onto their operations. Another involved complexities in teleoperation, the process of attempting to manage a robot from afar. The third required a system to address how data in motion, data at rest and computations could be stored, organized and queried.
In each case, it is essential for the robots to communicate in the most seamless, efficient and direct way possible. With this goal in mind, Zenoh improved R2X communication in a number of ways. The first was to overcome “connectivity islands” or areas where machines running on different network types fail to interact effectively with one another.
“Imagine that you have multiple robots that need to coordinate,” Corsaro says. “Most of the time, if they are the same robots, we're working on a fixed network, meaning those that use a wire, and everything will work fine because the characteristics of the network are alike. But as soon as you move to a wireless network, the world changes. This is where some of the technology that used to work just great on a fixed network will completely blow your wireless network apart. That’s where Zenoh comes into play. It drops a little service, and all of a sudden, all of your robots can talk.”
With respect to teleoperation, Corsaro highlights another aspect of Zenoh that makes R2X communication easier than ever before. In this scenario, a farmer might be sitting under a lemon tree on his tablet, attempting to give instructions to the autonomous machines in his field. Normally, to accomplish this task, the farmer would deploy a service that’s on the cloud, even though, in this example, the robot is stationed directly across the field. Communicating in this way—from the farmer in the field to the cloud to the robot in the same field—is extremely inefficient and more than a little silly.
“With Zenoh, users have the ability to always communicate in the most direct manner,” Corsaro says. “So, if the farmer is sitting under the lemon tree and teleoperating his robot, then the communication will be direct if there is data network connectivity. If for some reason, there isn’t any data network connectivity, we will find the closest place where we can complete the communication. The implication of communicating directly is to ensure that the closer you are, the better experience you have.”
Perhaps one of Zenoh’s most impressive features is that, in addition to having a reach that spans from the microcontroller to the data center, the protocol can also unify data in motion, data at rest and computations. Data in motion concerns the information collected with a robot is working. Data at rest is the information that is stored. Computation has to do with analyzing, synthesizing or evaluating the data. Zenoh manages all three in a way that allows a user to access the data they need when they need it. To understand how this is possible, it’s helpful to return to the example of the nervous system.
“The nervous system distributes information, but it also stores information on the periphery because otherwise, our reaction time would never be that fast,” Corsaro says. “The information gets stored in the memory. The body, then, is a good example of how this system should work. The way we design our system today, however, assumes that we will store all the information in the brain. If humans were designed the way we design our system, we would always burn our hand on the stove. So, why do we design the system as we do? Well, because the technologies that are available make it easy for us to store data in a single place. As soon as we try to store data in multiple places, we need to know the location of the data to get what we need.”
The Zenoh protocol works by crawling the entire network to find the location of where the information is stores and retrieve exactly what that the farmer needs. This enables the data to be stored locally on the robot itself, instead of sending it all to the cloud. It all comes back to efficient communication. Retrieving data from the cloud requires a lot more energy than finding it inside the robot’s memory system.
ZettaScale develops its solutions with energy efficiency, safety and security in mind.
While systems that foster direct communication may not seem like a huge deal, they are an essential part of minimizing energy consumption. The topic of energy and resource usage has become a prevalent one in the agricultural industry and around the world. ZettaScale builds its products with this conversation in mind.
“In physics, the thing that takes the most energy is not computation—it’s communication,” Corsaro says. “So, even if we have a super green data center that is very efficient, when I’m sitting in my living room and controlling my thermostat a few meters away from me, I’m still crossing a large portion of the internet to get to the data center and back. Nobody is really counting that energy right now. But the number of connected devices is growing exponentially, and if we continue on that path, the big elephant in the room will become figuring out precisely what portion of energy is consumed by communication.”
ZettaScale prefers to be ahead of the game. In addition to building solutions with energy consumption in mind, the company also prioritizes safety and security. Aside from the safety certifications ZettaScale is pursuing for Cyclone DDS and Zenoh, it places a high importance on using memory- and concurrency-safe programming languages.
According to a recent memo from the National Security Agency (NSA), “Microsoft and Google have each stated that software memory safety issues are behind around 70 percent of their vulnerabilities.” Currently, Zenoh is developed using RUST, one of the languages that is designed to protect against such vulnerabilities.
“We are very cautious about security,” Corsaro says. “From our perspective, memory- and concurrency-safe programming languages are essential for anything that communicates on the internet because as soon as you have something that’s communicating, there’s an open door where someone can go and try to sneak things through.”
Corsaro is aware that despite his company’s efforts to be ahead of the curve, users will continue to encounter new and interesting challenges. If it’s not a potential security threat, it’s the desire to adapt computer vision to account for the intensity of the sun rays based on where the farmer wants to send his machines (an actual client request ZettaScale has fielded). The company vows to continue developing solutions that rise to meet their customers’ unique needs.
“We are a high-tech company, and we have a bunch of super passionate employees that really like to solve the problems other people tend to ignore,” Corsaro says. “We strive to create solutions that address these issues, which makes the user much happier.”