In addition to the rubber balls, there’s a nice physical interface on the side for adjusting volume and pairing multiple Mix speakers together if you have multiple on hand (I was only sent the single mono speaker). Setup involves installing the Soundboks app, pairing to the speaker via Bluetooth on your phone, and picking whatever you want to play. It’s all quick and painless, especially for my first-time pairing with a Samsung Galaxy S24 Ultra.

Otherwise, it’s all very pro audio. Everything reminds me very much of the Peavey PA system I have in my music rehearsal space. The top of the speaker features a built-in carrying handle and a place for a strap (an accessory you have to buy aftermarket, or you can fasten it with any strap you have that fits through the hole). There are also top-hat mounts for the speakers to slide onto traditional PA pole stands, if you wanted to use them in that way at a party or event.

The grill is replaceable, as is the massive internal battery, which means that these things are pretty much indestructible as long as the amp and speakers themselves still work—the battery is the weak point of most portable speakers in 2026.

I bounced it around my yard, dropped it off my patio, and generally beat the crap out of it during my two-week testing period, and the thing just needed a little wipe down and a charge when it ran out of juice. The claimed 40 hours of battery at reasonable volume is accurate, but you’ll get about eight hours at max volume (which is very good for the category). If you need to bring some walk-out music to your kid’s all-day Little League tournament, this a great way to go.

Big Sound

Soundboks calls this speaker midsize, but at 21.4 pounds and the size of a medium-size cooler, I’d still call it a large speaker. That said, the size doesn’t make it any less portable than competitors from JBL and others; you still need a car or cargo ebike to take one of these with you, so what’s a couple inches here or there? The fact that this is a rectangle actually makes it easier to strap down than many others, especially with the holes for the strap and the built-in handle to tie down through.

Where the X10 Pro Omni had rotating mop pads, the rolling mop pad on the Omni C28 continuously self-cleans to prevent spreading dirt or grime to other parts of the house. Both apply downward pressure, but neither can spot dirtier places on their own as pricier, AI-powered robot vacuums will. Still, I was happy to see that it was able to scrub away some of the large dirt smudges in my entryway, though it didn’t get all of them. It also didn’t manage to scrub away all of the cherry juice I intentionally spilled in my routine mess setup for robot vacuum testing, even after sending the vacuum to do a second mopping job on one of the spots.

Still, the Omni C28 was able to raise its roller mop high enough when it switched from mopping my floors to vacuuming my living room rug that there was no hint of dampness anywhere. The older X10 did get my colleague Adrienne So’s carpet wet, but it didn’t get mine wet, though my carpet is a fairly low pile. It did a fine job vacuuming the carpet, though I could tell the difference in suction between this and more powerful vacuums I’ve tested.

The base station is nice and compact, and includes drying fans to dry off the roller mop. That does mean there’s a gentle fan noise in the background for a couple of hours after you use this robot vacuum, which was more annoying than I expected, but you could easily place this vacuum’s base station in a less central spot in your home so you don’t hear it. You could also set up a schedule for the vacuum to run in the morning and finish its drying job before you get home.

Multi-Floor Madness

Photograph: Nena Farrell

My favorite feature on the Omni C28 is that, even at this price point, it can still learn multiple maps. While it can’t climb up stairs, you can move it around your home and switch the maps in the app to the floor you’ve relocated to. This isn’t new for Eufy, as the older affordable model can do that too, but it’s nice to see the feature maintained when I’ve tried more expensive robot vacuums that don’t include it. It’s pretty simple to use; you’ll go to the maps, select “make a new map,” and then activate the robot to map. Once the map is made, you’ll switch to that map from the little map icon on the right side, which will label them with numbers in the order you created them.

Inside a giant autonomous warehouse, hundreds of robots dart down aisles as they collect and distribute items to fulfill a steady stream of customer orders. In this busy environment, even small traffic jams or minor collisions can snowball into massive slowdowns.

To avoid such an avalanche of inefficiencies, researchers from MIT and the tech firm Symbotic developed a new method that automatically keeps a fleet of robots moving smoothly. Their method learns which robots should go first at each moment, based on how congestion is forming, and adapts to prioritize robots that are about to get stuck. In this way, the system can reroute robots in advance to avoid bottlenecks.

The hybrid system utilizes deep reinforcement learning, a powerful artificial intelligence method for solving complex problems, to figure out which robots should be prioritized. Then, a fast and reliable planning algorithm feeds instructions to the robots, enabling them to respond rapidly in constantly changing conditions.

In simulations inspired by actual e-commerce warehouse layouts, this new approach achieved about a 25 percent gain in throughput over other methods. Importantly, the system can quickly adapt to new environments with different quantities of robots or varied warehouse layouts.

“There are a lot of decision-making problems in manufacturing and logistics where companies rely on algorithms designed by human experts. But we have shown that, with the power of deep reinforcement learning, we can achieve super-human performance. This is a very promising approach, because in these giant warehouses even a 2 or 3 percent increase in throughput can have a huge impact,” says Han Zheng, a graduate student in the Laboratory for Information and Decision Systems (LIDS) at MIT and lead author of a paper on this new approach.

Zheng is joined on the paper by Yining Ma, a LIDS postdoc; Brandon Araki and Jingkai Chen of Symbotic; and senior author Cathy Wu, the Class of 1954 Career Development Associate Professor in Civil and Environmental Engineering (CEE) and the Institute for Data, Systems, and Society (IDSS) at MIT, and a member of LIDS. The research appears today in the Journal of Artificial Intelligence Research.

Rerouting robots

Coordinating hundreds of robots in an e-commerce warehouse simultaneously is no easy task.

The problem is especially complicated because the warehouse is a dynamic environment, and robots continually receive new tasks after reaching their goals. They need to be rapidly redirected as they leave and enter the warehouse floor.

Companies often leverage algorithms written by human experts to determine where and when robots should move to maximize the number of packages they can handle.

But if there is congestion or a collision, a firm may have no choice but to shut down the entire warehouse for hours to manually sort the problem out.

“In this setting, we don’t have an exact prediction of the future. We only know what the future might hold, in terms of the packages that come in or the distribution of future orders. The planning system needs to be adaptive to these changes as the warehouse operations go on,” Zheng says.

The MIT researchers achieved this adaptability using machine learning. They began by designing a neural network model to take observations of the warehouse environment and decide how to prioritize the robots. They train this model using deep reinforcement learning, a trial-and-error method in which the model learns to control robots in simulations that mimic actual warehouses. The model is rewarded for making decisions that increase overall throughput while avoiding conflicts.

Over time, the neural network learns to coordinate many robots efficiently.

“By interacting with simulations inspired by real warehouse layouts, our system receives feedback that we use to make its decision-making more intelligent. The trained neural network can then adapt to warehouses with different layouts,” Zheng explains.

It is designed to capture the long-term constraints and obstacles in each robot’s path, while also considering dynamic interactions between robots as they move through the warehouse.

By predicting current and future robot interactions, the model plans to avoid congestion before it happens.

After the neural network decides which robots should receive priority, the system employs a tried-and-true planning algorithm to tell each robot how to move from one point to another. This efficient algorithm helps the robots react quickly in the changing warehouse environment.

This combination of methods is key.

“This hybrid approach builds on my group’s work on how to achieve the best of both worlds between machine learning and classical optimization methods. Pure machine-learning methods still struggle to solve complex optimization problems, and yet it is extremely time- and labor-intensive for human experts to design effective methods. But together, using expert-designed methods the right way can tremendously simplify the machine learning task,” says Wu.

Overcoming complexity

Once the researchers trained the neural network, they tested the system in simulated warehouses that were different than those it had seen during training. Since industrial simulations were too inefficient for this complex problem, the researchers designed their own environments to mimic what happens in actual warehouses.

On average, their hybrid learning-based approach achieved 25 percent greater throughput than traditional algorithms as well as a random search method, in terms of number of packages delivered per robot. Their approach could also generate feasible robot path plans that overcame congestion caused by traditional methods.

“Especially when the density of robots in the warehouse goes up, the complexity scales exponentially, and these traditional methods quickly start to break down. In these environments, our method is much more efficient,” Zheng says.

While their system is still far away from real-world deployment, these demonstrations highlight the feasibility and benefits of using a machine learning-guided approach in warehouse automation.

In the future, the researchers want to include task assignments in the problem formulation, since determining which robot will complete each task impacts congestion. They also plan to scale up their system to larger warehouses with thousands of robots.

This research was funded by Symbotic.