George Mason U. researchers enable robots to intelligently navigate challenging terrain

George Mason U. researchers enable robots to intelligently navigate challenging terrain

By Scott Simmie


Picture this: You’re out for a drive and in a hurry to reach your destination.

At first, the road is clear and dry. You’ve got great traction and things are going smoothly. But then the road turns to gravel, with twists and turns along the way. You know your vehicle well, and have navigated such terrain before.

And so, instinctually, you slow the vehicle to navigate the more challenging conditions. By doing so, you avoid slipping on the gravel. Your experience with driving, and in detecting the conditions, has saved you from a potential mishap. Yes, you slowed down a bit. But you’ll speed up again when the conditions improve. The same scenario could apply to driving on grass, ice – or even just a hairpin corner on a dry paved road.

For human beings, especially those with years of driving experience, such adjustments are second-nature. We have learned from experience, and we know the limitations of our vehicles. We see and instantly recognize potentially hazardous conditions – and we react.

But what about if you’re a robot? Particularly, a robot that wants to reach a destination at the maximum safe speed?

That’s the crux of fascinating research taking place at George Mason University: Building robots that are taught – and can subsequently teach themselves – how to adapt to changing terrain to ensure stable travel at the maximum safe speed.

It’s very cool research, with really positive implications.

Below: You don’t want this happening on a critical mission…

George Mason Xuesu Xiao Hunter SE



Those are the initials of Dr. Xuesu Xiao, an Assistant Professor at George Mason University. He holds a PhD in Computer Science, and runs a lab that plays off his initials, called the RobotiXX Lab. Here’s a snippet of the description from his website:

“At RobotiXX lab, researchers (XX-Men) and robots (XX-Bots) perform robotics research at the intersection of motion planning and machine learning with a specific focus on robustly deployable field robotics. Our research goal is to develop highly capable and intelligent mobile robots that are robustly deployable in the real world with minimal human supervision.”

We spoke with Dr. Xiao about this work.

It turns out he’s particularly interested in making robots that are particularly useful to First Responders, and carrying out those dull, dirty and dangerous tasks. Speed in such situations can be critical, but comes with its own set of challenges. A robot that makes too sharp a turn at speed on a high friction surface can easily roll over – effectively becoming useless in its task. Plus, there are the difficulties previously flagged with other terrains.

This area of “motion planning” fascinates Dr. Xiao. Specifically, how to take robots beyond traditional motion planning and enable them to identify and adapt to changing conditions. And that involves machine vision and machine learning.

“Most motion planners used in existing robots are classical methods,” he says. “What we want to do is embed machine learning techniques to make those classical motion planners more intelligent. That means I want the robots to not only plan their own motion, but also learn from their own past experiences.”

In other words, he and his students have been focussing on pushing robots to develop capabilities that surpass the instructions and algorithms a roboticist might traditionally program.

“So they’re not just executing what has been programmed by their designers, right? I want them to  improve on their own, utilising all the different sources of information they can get while working in the field.”




The RobotiXX Lab has chosen the Hunter SE from AgileX as its core platform for this work. That platform was supplied by InDro Robotics, and modified with the InDro Commander module. That module enables communication over 5G (and 4G) networks, enabling high speed data throughput. It comes complete with multiple USB slots and the Robot Operating System (ROS) library onboard, enabling the easy addition (or removal) of multiple sensors and other modifications. It also has a remote dashboard for controlling missions, plotting waypoints, etc.

Dr. Xiao was interested in this platform for a specific reason.

“The main reason is it is because it’s high speed, with a top speed of 4.8m per second. For a one-fifth/one-sixth scale vehicle that is a very, very high speed. And we want to study what will happen when you are executing a turn, for example, while driving very quickly.”

As noted previously, people with driving experience instinctively get it. They know how to react.

“Humans have a pretty good grasp on what terrain means,” he says. “Rocky terrain means things will get bumpy, grass can impede a motion, and if you’re driving on a high-friction surface you can’t turn sharply at speed. We understand these phenomenon. The problem is, robots don’t.”

So how can we teach robots to be more human in their ability to navigate and adjust to such terrains – and to learn from their mistakes?

As you’ll see in the diagram below, it gets *very* technical. But we’ll do our best to explain.

George Mason Hunter Xuesu Xiao



The basics here are pretty clear, says Dr. Xiao.

“We want to teach the robots to know the consequences of taking some aggressive maneuvers at different speeds on different terrains. If you drive very quickly while the friction between your tires and the ground is high, taking a very sharp turn will actually cause the vehicle to roll over – and there’s no way the robot by itself will be able to recover from it, right? So the whole idea of the paper is trying to enable robots to understand all these consequences; to make them ‘competence aware.'”

The paper Dr. Xiao is referring to has been submitted for scientific publication. It’s pretty meaty, and is intended for engineers/roboticists. It’s authored by Dr. Xiao and researchers Anuj Pokhrel, Mohammad Nazeri, and Aniket Datar. It’s entitled: CAHSOR: Competence-Aware High-Speed Off-Road Ground Navigation in SE(3).

That SE(3) term is used to describe how objects can move and rotate in 3D space. Technically, it stands for Special Euclidean group in three dimensions. It refers to keeping track of an object in 3D space – including position and orientation.

We’ll get to more of the paper in a minute, but we asked Dr. Xiao to give us some help understanding what the team did to achieve these results. Was it just coding? Or were there some hardware adjustments as well?

Turns out, there were both. Yes, there was plenty of complex coding. There was also the addition of an RTK GPS unit so that the robot’s position in space could be measured as accurately as possible. Because the team soon discovered that intense vibration over rough surfaces could loosen components, threadlock was used to keep things tightly in place.

But, as you might have guessed, machine vision and machine learning are a big part of this whole process. The robot needs to identify the terrain in order to know how to react.

We asked Dr. Xiao if an external data library was used and imported for the project. The answer? “No.”

“There’s no dataset out there that includes all these different basic catastrophic consequences when you’re doing aggressive maneuvers. So all the data we used to train the robot and to train our machine learning algorithms were all collected by ourselves.”




As part of the training process, the Hunter SE was driven over all manner of demanding terrain.

“We actually bumped it through very large rocks many times and also slid it all over the place,” he says. “We actually rolled the vehicle over entirely many times. This was all very important for us to collect some data so that it learns to not do that in the future, right?”
And while the cameras and machine vision were instrumental in determining what terrain was coming up, the role of the robot’s Inertial Measurement Unit was also key.

“It’s actually multi-modal perception, and vision is just part of it. So we are looking at the terrain using camera images and we are also using our IMU. Those inertial measurement unit readings  sense the acceleration and the angular velocities of the robot so that it can better respond,” he says.

“Because ultimately it’s not only about the visual appearance of the terrain, it is also about how you drive on it, how you feel it.”




Well, they’re impressive.

The full details are outlined in this paper, but here’s the headline: Regardless of whether the robot was operating autonomously heading to defined waypoints, or whether a human was controlling it, there was a significant reduction in incidents (slips, slides, rollovers etc.) with only a small reduction in overall speed.

Specifically, “CAHSOR (Competence-Aware High-Speed Off-Road Ground Navigation) can efficiently reduce vehicle instability by 62% while only compromising 8.6% average speed with the help of TRON (visual and inertial Terrain Representation for Off-road Navigation).”

That’s a tremendous reduction in instability – meaning the likelihood that these robots will reach their destination without incident is greatly improved. Think of the implications for a First Responder application, where without this system a critical vehicle rushing to a scene carrying medical supplies – or even simply for situational awareness – might roll over and be rendered useless. The slight reduction in speed is a small price to pay for greatly enhancing the odds of an incident-free mission.

“Without using our method, a robot will just blindly go very aggressively over every single terrain – while risking rolling over, bumps and vibrations on rocks, maybe even sliding and rolling off a cliff.”

What’s more, these robots continue to learn with each and every mission. They can also share data with each other, so that the experience of one machine can be shared with many. Dr. Xiao also says the learnings from this project, which began in January of 2023, can also be applied to marine and even aerial robots.

For the moment, though, the emphasis has been fully on the ground. And there can be no question this research has profound and positive implications for First Responders (and others) using robots in mission-critical situations.

Below: The Hunter SE gets put through its paces. (All images courtesy of Dr. Xiao.)

Hunter SE George Mason Xuesu Xiao



We’re tremendously impressed with the work being carried out by Dr. Xiao and his team at George Mason University. We’re also honoured to have played a small role in supplying the Hunter SE, InDro Commander, as well as occasional support as the project progressed.

“The use of robotics by First Responders is growing rapidly,” says InDro Robotics CEO Philip Reece. “Improving their ability to reach destinations safely on mission-critical deployments is extremely important work – and the data results are truly impressive.

“We are hopeful the work of Dr. Xiao and his team are adopted in future beyond research and into real-world applications. There’s clearly a need for this solution.”

If your institution or R&D facility is interested in learning more how InDro’s stable of robots (and there are many), please reach out to us here.

New LIMO Pro, ROS2 models bring advanced abilities to R&D

New LIMO Pro, ROS2 models bring advanced abilities to R&D

By Scott Simmie


There’s a new robot in town. Actually, there are two of them.

They’re small but mighty. In fact, numerous universities and robotics labs already use their predecessor for high-level research. That original robot, the AgileX LIMO, was a game-changer when it came to an affordable and flexible platform. Boston University currently has a fleet of LIMOs running custom algorithms and simulations related to how real-world autonomous vehicles will interact in the Smart Cities of the future. (It’s really cool research, and you can read all about it here.)

That first LIMO was truly a ground-breaker – and remains an excellent R&D research platform. But now, AgileX has taken things further. Two new versions of LIMO offer advanced hardware, software, runtime – and capabilities.

Below: The original LIMO that started it all…

AgileX Limo Robot



Before we get into the significant changes incorporated into the new models, it’s worth looking at some of the strong features common to all members of the LIMO family. For starters, each LIMO has four steering modes: Omnidirectional steering, tracked steering, Ackerman and four-wheel differential.

All LIMOs are equipped with obstacle detection. Multiple onboard sensors can pick up on the size, distance and location of obstacles, allowing the robot to navigate its environment without conflict. The newer LIMO models have significant enhancements here, which we’ll explore in a moment.

Despite their relatively small size – so small an untrained eye might potentially mistake them for a toy – the LIMOs feature a robust, all-metal build and powerful motor. They also feature powerful onboard EDGE computing suitable to pretty much any R&D requirements.




There are really two main categories of users, with the first being those in the educational field.

“This is a great tool for anyone looking to learn ROS, because they can do all of the advanced concepts – obstacle detection, SLAM, teleoperation, to name a few,” explains Luke Corbeth, Head of InDro’s R&D Sales Division.

“And we make that really simple through our improved documentation. We’ve basically built a course around it, so it can be used for teaching students.”

The other main group of users, of course, are on the research side of things.

“It’s almost always used in labs for multi-agent systems or multiple robot projects. Because it’s multi-modal, when you’re doing a multi-agent system it can be homogenous or heterogeneous, meaning you use different steering in different robots simultaneously.”

Dimensions of all versions of LIMO are identical, as seen below.

AgileX LIMO Robot



The original LIMO is still a great robot – and is currently in use by many universities. But AgileX didn’t rest on its laurels. In response to the availability of new technologies – along with a wish-list from existing clients – the company has taken things further with its new LIMO PRO and the LIMO ROS2.

“Obviously as research in autonomy advances, so do the computational requirements,” explains Corbeth. “So it’s very important as a robotics manufacturer to stay ahead of the curve so that the hardware meets up with the current research requirements of the day.”

To that end, the new versions feature upgrades on computational power, sensors and run-time.

“The big difference is in compute – we’re moving from the Jetson Xavier to the Jetson Orin Nano on the LIMO PRO and the INTEL NUC on the LIMO ROS2. Both of these are actually massive upgrades.”

The Orin Nano is a very powerful EDGE computer. That power translates into more stable multi-sensor data fusion and speed with SLAM (Simultaneous Localization and Mapping) processing.

Speaking of SLAM, the LIMO PRO and LIMO ROS2 come with a new LiDAR unit. While the original LIMO used the very capable EAI X2L unit, the new versions come with the EAI T-mini Pro.

“Plus, the battery in the new unit goes from an hour of run-time to 2-1/2 hours – with a standby of four hours,” adds Corbeth.




Not surprisingly, the two new versions also feature some software upgrades. The LIMO PRO and ROS2 versions come with Ubuntu 20.04 (the original LIMO runs version 18.04). In terms of ROS (Robot Operating System) libraries, the first generation LIMO is outfitted with ROS1 Melodic. The LIMO PRO features both ROS1 Noetic and ROS2 Foxy. The LIMO ROS2 has ROS2 Humble onboard.

Already have some of the first-generation LIMOs in your lab? No problem.

“The new models can co-exist with the original LIMO,” says Corbeth. “And if the computing demands are higher than previous applications, it makes sense of have a blend of models.”

The graphic below outlines the feature sets of the three models:

LIMO Robot Canada



InDro Robotics has a lot of clients who have put the original LIMO to use in labs and educational institutes across North America. Boston University has a very large fleet of LIMOs deployed – hard at work on multiple research projects related to Smart Cities and autonomous vehicles. They’ve proven to be a robust, cost-effective tool for high-level research.

And now, with the fresh release of LIMO PRO and LIMO ROS2, there are two more affordable options.

“This is a significant development for anyone looking to expand their current fleet of LIMOS, as well as those who have been waiting in the wings for an upgrade,” says Corbeth. “These are incredibly powerful and versatile robots/research tools, with the added bonus that the entire line is very affordable.”

If you’re interested in learning more, InDro Robotics is the exclusive distributor of AgileX in Canada, as well as a distributor for all of North America. We have built excellent documentation and manuals to assist users ranging from beginning to expert – and all of that added value and support comes with every purchase made through InDro.

For more information from someone who really knows their stuff, contact Luke Corbeth here.

uPenn robotics team cleans up at SICK LiDAR competition

uPenn robotics team cleans up at SICK LiDAR competition

By Scott Simmie


There’s nothing we like more than success stories – especially when technology is involved.

So we’re pleased to share news that a team of bright young engineers from the University of Pennsylvania were the winners of a prestigious competition sponsored by SICK, the German-based manufacturer of LiDAR sensors and industrial process automation technology.

The competition is called the SICK TiM $10K Challenge. The competition involves finding innovative new uses for the company’s TiM-P 2D LiDAR sensor. Laser-based LiDAR sensors scan the surrounding environment in real-time, producing highly accurate point clouds/maps. Paired with machine vision and AI, LiDAR can be used to detect objects – and even avoid them.

And that’s a pretty handy feature if your robot happens to an autonomous garbage collector. We asked Sharon Shaji, one of five UPenn team members (all of whom earned their Masters in Robotics this year), for the micro-elevator pitch:

“It’s an autonomous waste collection robot that can be used specifically for cleaning outdoor spaces,” she says.

And though autonomous, it obviously didn’t build itself.

Below: Members of the team during work on the project.

uPenn Sauberbot



When SICK announced the contest, it had a very simple criteria: “The teams will be challenged to solve a problem, create a solution, and bring a new application that utilizes the SICK scanner in any industry.”

SICK received applications from universities across the United States. It then whittled those down to 20 submissions it felt had real potential, and supplied those teams with the TiM-P 270 LiDAR sensor free of charge.

Five students affiliated with UPenn’s prestigious General Robotics, Automation, Sensing and Perception Laboratory, or GRASP Lab, put in a team application. It was one of three GRASP lab teams that would receive sensors from SICK.

That Lab is described here as “an interdisciplinary academic and research center within the School of Engineering and Applied Sciences at the University of Pennsylvania. Founded in 1979, the GRASP Lab is a premier robotics incubator that fosters collaboration between students, research staff and faculty focusing on fundamental research in vision, perception, control systems, automation, and machine learning.”

Before we get to building the robot, how do you go about building a team? Do you just put smart people together – or is there a strategy? In this case, there was.

“One thing we all kept in mind when we were looking for teammates was that we wanted someone from every field of engineering,” explains Shaji. In other words, a multidisciplinary team.

“So we have people from the mechanical engineering background, electrical engineering background, computer science background, software background. We were easily able to delegate work to every person. I think that was important in the success of the product. And we all knew each other, so it was like working with best friends.”




And how did the idea come about?

Well, says the team (all five of whom hopped on a video call with InDro Robotics), they noticed a problem in need of a solution. Quite frequently on campus – and particularly after events – they’d noticed that the green space was littered. Cans, bottles, wrappers – you name it.

They also noticed that crews would be dispatched to clean everything up. And while that did get the job done, it wasn’t perhaps the most efficient way of tackling the problem. Nor was it glamorous work. It was arguably a dirty and dull job – one of the perfect types of tasks for a robot to take on.

“Large groups of people were coming in and manually picking up this litter,” says Shaji.

“And we realised that automation was the right way to solve that problem. It’s unhygienic, there are sanitation concerns, and physically exhausting. Robots don’t get tired, they don’t get exhausted…we thought this was the best use-case and to move forward with.”

Below: Working on the mechanical side of things

uPenn SICK Sauberbot



You’d think, with engineers, the first step in this project would have been to kick around design concepts. But the team focussed initially on market research. Were there similar products out there already? Would there be a demand for such a device? How frequently were crews dispatched for these cleanups? How long, on average, does it take humans to carry out the task? How many people are generally involved? Those kinds of questions.

After that process, they began discussing the nuts and bolts. One of the big questions here was: How should the device go about collecting garbage? Specifically, how should it get the garbage off the ground?

“Cleaning outdoor spaces can vary, because outdoor spaces can vary,” says team member Aadith Kumar. “You might have sandy terrain, you might have open parks, you might have uneven terrain. And each of these pose their own problems. Having a vacuum system on a beach area isn’t going to work because you’re going to collect a lot of sand. The vision is to have a modular mechanism.”

A modular design means flexibility: Different pickup mechanisms would be swappable for specific environments without requiring an entirely new robot. A vacuum system might work well in one setting, a system with the ability to individually pick items of trash might work better somewhere else.

The team decided their initial prototype should focus on open park space. And once that decision was made, it became clear that a brush mechanism, which would sweep the garbage from the grass into a collection box, would be the best solution for this initial iteration.

“We considered vacuum, we considered picking it up, we considered targeted suction,” says Kumar. “But at the end of the day, for economics, it needed to be efficient, fast, nothing too complicated. And the brush mechanism is tried and tested.”

Below: Work on the brush mechanism



uPenn SICK Sauberbot



The team decided to call its robot the SauberBOT. “Sauber” is the German word for “clean”. But that sweeping brush mechanism would be just one part of the puzzle. Other areas to be tackled included:

  • Depth perception camera for identifying trash to be picked up
  • LiDAR programmed so that obstacles, including people, could be avoided
  • Autonomy within a geofenced location – ie, the boundaries of the park to be cleaned

There was more, of course, but one of the most important pieces of the puzzle was the robotic platform itself: The means of locomotion. And that’s where InDro Robotics comes in.




Some team members had met InDro Account Executive Luke Corbeth earlier in the year, at the IEEE International Conference on Robotics and Automation, held in Philadelphia in 2022. Corbeth had some robotic platforms from AgileX – which InDro distributes in North America – at the show. At the time the conference took place, the SICK competition wasn’t yet underway. But the students remembered Corbeth – and vice versa.

Once the team formed and entered the contest, discussions with InDro began around potential platforms.

The team was initially leaning toward the AgileX Bunker – a really tough platform that operates with treads, much like a tank. At first glance, those treads seemed like the ideal form of locomotion because they can operate on many different surfaces.

But Luke steered them in a different direction, toward the (less-expensive) Scout 2.0.

“He was the one who suggested the Scout 2.0,” says Udayagiri.

“We actually were thinking of going for the Bunker – but he understood that for our use-case the Scout 2.0 was a better robot. And it was very easy to work with the Scout.”

Corbeth also passed along the metal box that houses the InDro Commander. This enabled the team to save more time (and potential hassle) by housing all of their internal components in an IP-rated enclosure.

“I wanted to help them protect their hardware in an outdoor environment,” he says. “They had a tight budget, and UPenn is a pretty prominent robotics program in the US.”

But buying from InDro begs the question: Why not build their own? A team of five roboticists would surely be able to design and build something like that, right? Well, yes. But they knew they were going to have plenty of work on their own without having to build something from scratch. Taking this on would divert them from their core R&D tasks.

“We knew we would do it in a month or two,” says the team’s Rithwik Udayagiri. “But that would have left us with less time for market research and actually integrating our product, which is the pickup mechanism. We would have been spending too much time on building a platform. So that’s why we went with a standalone platform.”

It took a little longer than planned to get the recently released Scout 2.0 in the hands of the UPenn team. But because of communication with Luke (along with the InDro-supplied use of the Gazebo robot simulation platform), the team was able to quickly integrate the rest of the system with Scout 2.0 soon after it arrived.

“The entire project was ROS-based (Robot Operating System software), and they used our simulation tools, mainly Gazebo, to start working on autonomy,” explains Corbeth. “Even though it took time to get them the unit, they were ready to integrate their tech and get it out in the field very quickly. That was the one thing that blew me away was how quickly they put it together.”

It wasn’t long before SauberBOT was a reality. The team produced a video for its final submission to SICK. The SauberBOT team took first place, winning $10,000 plus an upcoming trip to Germany, where they’ll visit SICK headquarters.

Oh, and SauberBOT? The team says it cleans three times more quickly than using a typical human crew. 

Here’s the video.




Team SauberBOT knows some people are wary of robots. Some believe they will simply replace human positions and put people out of work.

That’s not the view of these engineers. They see SauberBOT – and other machines like it – as a way of helping to relieve people from boring, physically demanding and even dangerous tasks. They also point out that there’s a labour shortage, particularly in this sector.

“The cleaning industry is understaffed,” reads a note sent by the team. “We choose to introduce automation to the repetitive and mundane aspects of the cleaning industry in an attempt do the tasks that there aren’t enough humans to do.”
And what about potential jobs losses?
“We intend to make robots that aren’t aimed to replace humans,” they write.
“We want to equip the cleaning staff with the tools to handle the mundane part of cleaning outdoor spaces and therefore allow the workforce to target their attention to the more nuanced parts of cleaning which demand human attention.”
In other words, think of SauberBOT as a co-operative robot meant to assist but not replace humans. These are sometimes called “co-bots.” 
Below: Testing out the SauberBOT in the field
UPenn SICK SauberBOT



We’re obviously pleased to have played a small role in the success of the UPenn team. And while we often service very large clients – including building products on contract for some global tech giants – there’s a unique satisfaction that comes from this kind of relationship.

“It’s very gratifying,” says Corbeth. “In fact, it’s the essence of what I try to do: Enable others to build really cool robots.”

The SauberBOT is indeed pretty cool. And InDro will be keeping an eye on what these young engineers do next.

“The engineering grads of today are tomorrow’s startup CEOs and CTOs,” says InDro Robotics Founder/CEO Philip Reece.

“We love seeing this kind of entrepreneurial spirit, where great ideas and skills lead to the development of new products and processes. In a way, it’s similar to what InDro does on a larger scale. Well done, Team SauberBOT – there’s plenty of potential here for a product down the road.”

If you’ve got a project that could use a robotic platform – or any other engineering challenge that taps into InDro’s expertise with ground robots, drones and remote teleoperations – feel free to get in touch with Luke Corbeth here.

There’s a new robot in town: Meet LIMO

There’s a new robot in town: Meet LIMO

Even in the world of robotics, good things often come in small packages. And this is especially true when it comes to Limo, a new AgileX platform perfect for students and those carrying out R&D work. Limo is small but mighty, with the same kind of technology you’ll find in much larger devices (it weighs but 4.2 kg). The robot runs on the open source Robot Operating System (ROS) software, and comes with both the original ROS1 and ROS2 software libraries. This allows users to customize the robot for different tasks.

It ships with an impressive display of hardware and capabilities right out of the box, including:

  • An NVIDIA Jetson Nano, capable of remote teleoperation over 4G
  • An EAI X2L LiDAR unit
  • Stereo camera

This affordable machine is capable of autonomous missions, including mapping new surroundings via Simultaneous Localisation and Mapping (SLAM). It also comes with multiple modes for locomotion. You’ll see details of this in the left-hand graphic below. It’s also scalable. Want to add other sensors? There are four USB Serial Ports onboard.

This kind of flexibility in a small package is pretty amazing.


How Limo came about


We were curious to learn more about Limo, so we contacted AgileX’s Brandy Xue. Until recently, Brandy was leading the company’s Global Sales and Marketing department. In March of 2022, she switched to the new AgileX subsidiary, Mammotion Tech – which focuses on consumer outdoor robots like autonomous lawnmowers.

We started with a simple question. Who would be interested in buying Limo? Would it be primarily students? Researchers? Developers?

Her answer was simple: “Limo is for everybody,” she said. She then went on to explain why.

Many students, particularly in Southeast Asia, are now delving into coding, robotics – and even AI – while in high school. It’s been a trend in South Korea, and is being seen more and more in China. In fact, says Xue, the Chinese government has been encouraging hands-on high-tech training in high school to prepare people for the workforce.

“The policies in China supporting robotics education are growing,” she says. “And in South Korea, students are working on AI and Machine Learning in high school.”


Not just students


So AgileX knew there was an educational market for a product like this. But it also felt that researchers in the R&D world could also benefit from a robot with full-scale capabilities in an affordable, smaller-scale package. Having everything integrated out of the box saves a lot of groundwork. Plus, many smaller companies don’t have the need (or the budget) for a larger machine.

“If they want to build a robot, they have to buy a robot here, a sensor there, then write the code to make it move. It’s too complicated,” she says.

“And most people don’t know what to buy, or don’t know how to write the code at the beginning. So why don’t we do this to make it easier for the developer to build a robot? It’s a really cost-effective solution.”And so they did. It also didn’t hurt that the company’s CEO, JD Wei, ran the impressive Robomaster division at DJI. Annual Robomaster competitions pit robots built by the best and brightest teams of engineers against one another. DJI has also hired a significant number of engineers through the program, which has grown since its inception to become more global in nature.

If you’re unfamiliar with Robomaster, check out the video below. It’s worth watching, as it also gives you a pretty good idea of the background JD Wei came from:


Simulation table


Because Limo is capable of autonomous movement, it can be purchased with an optional simulation table. That platform approximates a mini-city, complete with buildings, roads, stop signs, traffic lights – even a liftable gate arm, like you’d see at railroad crossings or when exiting a parking lot.

Limo can detect and act on its surroundings and can be programmed to take different actions depending on the environment. It can even use its onboard LiDAR to create a 3D, Virtual SLAM map of what it “sees” around it.

The complete package is covered in this AgileX video, which also highlights its multi-modal locomotion capabilities.


Powerful processor and more…


Limo comes equipped with enviable brains. It features the NVIDIA Jetson Nano processor for EDGE computing. The Jetson is a powerful tool for AI development, and NVIDIA’s JetPack SDK offers even more options for deep learning, computer vision and more. It’s also 4G-compatible for remote tele-operation.

InDro’s Head of Robotic solutions, Peter King, is impressed with the package – saying it offers students and developers an affordable solution for R&D and prototyping.

“Limo really fills a void in the marketplace, allowing schools, researchers, and even R&D companies with limited budgets access to a truly powerful and expandable platform,” says King.

Limo is also rugged. The body is metal, and the 4.2 kg device is capable of tackling inclines of 25°. You’ll see the rest of the specs here:




Limo, as you can see, can do a lot on its own. And it’s capable of doing much more in the hands of a skilled developer or a motivated student. Given that this SLAM-capable device comes with a LiDAR unit, stereo camera, the NVIDIA Jetson Nano, and an onboard 7″ touchscreen module, you’d rightly expect it to cost a significant amount.

It doesn’t. The Limo is $2900 US in its base, multi-modal form. The simulation table, which offers a head-start for those interested in autonomous operation in a city-like environment, is available for an additional $1,000 US. If you’re interested in seeing Limo, we’re happy to arrange for a remote demonstration. You can reach us here.



InDro’s Take


We’ve always been impressed with the AgileX products. They’re smartly engineered and very well-constructed. Our Sentinel teleoperated inspection robot is built on the AgileX Bunker platform, capable of operating in even the most unforgiving of environments. In a word: AgileX builds great stuff. And the flexible design of its products means many are destined for even greater things.

That doesn’t surprise us, given CEO JD Wei’s background running DJI’s Robomaster program.

“After he left DJI, he founded AgileX Robotics – and he’s always joking to himself,” laughs Xue. “He used to work in a company whose robots fly in the sky. Now he runs a company whose robots run on the road.”

And, with the Limo, in classrooms and R&D labs as well.