logo
Multimodal AI makes this robot vacuum cleaner smart enough to pick up your socks

Multimodal AI makes this robot vacuum cleaner smart enough to pick up your socks

As long ago as 1962, The Jetsons imagined the vacuum cleaner as more than a simple appliance. Rosie was a friendly, helpful, and intelligent presence in the household, helping the family tidy up after themselves. That vision is no longer some far-flung fantasy of the future. It's real and ready for mass production thanks to a specific development in AI.
This May, Roborock is launching Saros Z70, the world's first mass-produced robot vacuum cleaner with an OmniGrip intelligent, foldable robotic arm that can pick up your socks and put them back where they need to be.
What makes this possible is "multimodal AI" — an intelligence capable of making sense of data from multiple modalities, such as infrared cameras and sensors. Instead of relying on a single input such as a laser, it learns from a fusion of information from all sorts of data points.
In the case of the Saros Z70, it uses 3D point cloud data (similar to what's used for facial recognition) to see and measure the depth between objects in your home. It then continuously updates its map of every room in real time so the device can rapidly adapt to new furniture layouts and obstacles.
But as astonishing as this technology is, it didn't come from nowhere. The first big shift in this type of AI was when Roborock applied its StarSight Autonomous System, which is a navigation and obstacle avoidance system, so that robot vacuums could use a laser to navigate around the house.
This next iteration of that system uses a sampling frequency that's 21 times higher than traditional LDS solutions, scanning 21,600 sensor points to identify objects as small as two square centimeters, so it can not just vacuum but actively organize your home for you. The result is that if it comes across a piece of crumpled paper on your floor, it's intelligent enough to know it needs to be treated differently than a sock.
At the heart of this massive leap forward is a multimodal AI that's designed to ensure this generation of robot vacuum will continue to become even smarter, safer and friendlier than it already is.
"The Saros Z70 can help users grasp the future with intelligence," said Richard Chang, founder and CEO, Roborock. "It can be a user's home helper, companion, and assistant — whatever it needs to be."
The magic of AI powered by multimodal perception
Robotic arms are already widely used in all sorts of industrial settings. But for a robot vacuum to operate a 5-axis robotic arm in a home, it needs to be intelligent enough to navigate not just all your furniture but also whatever surprises your kids and pets might throw its way. That takes multimodal perception - an ability to see the home in a multitude of different ways and the intelligence to make sense of all that data.
The Saros Z70 accomplishes this by combining 3D Time-of-Flight (ToF) sensors and visual recognition to create a real-time 3D map of everything around it so it can detect objects, recognize spatial relationships between them, and intelligently adapt.
It uses RGB and infrared cameras in the arm and on its chassis to perceive depth more accurately than previous LIDAR systems. The lateral wall sensor extends the field of vision made possible by a single-point laser to eliminate blind spots so the vacuum can keep cleaning even when it needs to move alongside something like a cable on the floor. And its AI uses a range of sensors to make sure the foldable arm is only used in the safest possible ways.
For instance, it can clean areas that were previously obstructed and lift socks, small towels, tissue papers, and sandals. But its sensors also prevent potential issues.
The weight sensor in the arm prevents its gentle grip from lifting anything heavier than 300 grams, the anti-pinch and pressure sensors make sure you can interact with the arm safely, and the upward spatial sensor keeps the arm from unfurling if a child, cat, or coffee table is 45 cm above it.
The combination of all these different data sources ensures the device can dynamically react to the specific objects and obstacles in any given home. It needs multimodal AI to align and fuse all these different inputs into a unified view of reality that is greater than the sum of its parts.
Each one of these high-definition sensors provides valuable context, but it's the AI in the middle that can turn all that data into the intelligence needed to safely clean a home that's constantly changing with kids and pets running around.
A friendly robot that grows smarter all the time
Remarkably, all those sensors and cameras are packed into the thinnest Roborock yet, a mere 7.98 cm (3.14 inches) tall. So it's smart enough to tell the difference between a cat and a lost slipper, but it's also thin enough to clean under the sofa.
The upshot of all this intelligence is a robot vacuum that's capable of learning your habits automatically so it knows when to pick something up, where to put it and whether it needs to use suction or mopping for optimal results on all the different routes it takes.
The device can also be voice-activated and used to keep an eye on your pets, using its cameras to find them wherever they are in the house and even send you pictures.
It's a truly friendly assistant that is constantly becoming more intelligent the more it gets to know you. It adapts to your home layouts when they change, reacts to objects when they're out of place and is continuously updated over-the-air to detect new objects.
It already knows, for instance, not to pick up a pair of scissors while being smart enough to put a crumpled piece of paper in the bin.
All of which means that the vision of the multi-tasking robot tending to your home is no longer just a vision. There's already a friendly AI-powered robot you can talk to while it puts your socks away and it's called the Saros Z70.
Orange background

Try Our AI Features

Explore what Daily8 AI can do for you:

Comments

No comments yet...

Related Articles

Roborock S8 Max Ultra Is $500 Off, Amazon Pushes to Clear Stock Fast
Roborock S8 Max Ultra Is $500 Off, Amazon Pushes to Clear Stock Fast

Gizmodo

timea day ago

  • Gizmodo

Roborock S8 Max Ultra Is $500 Off, Amazon Pushes to Clear Stock Fast

Roborock represents the ultimate luxury in robot vacuum cleaners, and it offers a broad range of models that span from around $300 up to $2,000. It's all about discovering that ideal sweet spot between price and performance, and the S8 Max Ultra nails it. The model belongs to Roborock's premium 'S' line of robot vacuum cleaners which is known for delivering the utmost cleaning power and smart features. Amazon is now offering a huge discount of $500 off, bringing the cost down to $799 (from $1,299), which makes the offer hard to beat for a product in its category. See at Amazon Our Favorite Model Cleaning performance on the S8 Max Ultra is excellent thanks to its powerful 8,000Pa suction power as it effortlessly cleans dirt, dust, and debris on various floors. Mop is improved by dual vibration modules that scrub at 4,000 times per minute so it also works well on dry stains. Smart navigation is driven by the PreciSense LiDAR system and Reactive 3D Obstacle Avoidance so that the robot can create detailed maps and skillfully navigate around obstacles with 3D structured light tech. At the heart of the system is the improved RockDock Ultra, its very own docking station, which revolutionizes maintenance by automatically washing the mop with hot water at temperatures of up to 140°F. The hot water cleaning effortlessly breaks up stubborn stains and the dock intelligently senses levels of grime and adjusts wash times accordingly. It also blows hot air to dry the mop and prevent odors, in addition to supplying over three months' worth of detergent. The dock even takes care of dust collection, water refilling and self-cleaning of the base. The updated Roborock app also brings intelligent control to the next level with more sophisticated AI algorithms. Its SmartPlan automatically assigns cleaning intensities based on room type and dirt levels. Temporary Cleaning is also enabled in the app to handle interruptions and a special Pet Area Deep Cleaning mode is available to deeply clean pet accidents. Other app features like Smart Suggestions for no-go zones and detailed 3D mapping make it easy to personalize and optimize the cleaning experience. The Roborock S8 Max Ultra is a total cleaning solution with powerful suction, precise navigation, auto care, and smart app controls. Amazon's $500 discount today brings this premium model down to $799, and this sale is an offer you cannot refuse. See at Amazon

I Used This Robot Vacuum With a Mechanical Arm for Three Months but at $2,600 It's a Hard Sell
I Used This Robot Vacuum With a Mechanical Arm for Three Months but at $2,600 It's a Hard Sell

CNET

time3 days ago

  • CNET

I Used This Robot Vacuum With a Mechanical Arm for Three Months but at $2,600 It's a Hard Sell

The Saros Z70 is priced at $2,599, making it one of the most expensive robot vacuums I've seen. You can buy two robot vacuums for this price. The most notable feature is the OmniGrip mechanical arm designed to pick up and sort items. It also has strong vacuuming and mopping capabilities, excellent mapping and the ability to cross thresholds. The mechanical arm has several issues, including difficulties picking up objects, navigation problems and software bugs. There were also connectivity issues during setup. Cleaning my hardwood and carpet. Ajay Kumar/CNET Pros Good vacuuming on hardwood with 82.5% sand pickup Very low profile and fits under pretty much everything Quiet operation Works well in homes with pets Chassis lift prevents it from getting stuck most of the time Cons Dust bin is too small Water tank is too small Mopping performance isn't great Obstacle avoidance needs some improvement Expensive at $2,600 at full price Roborock I've been using the Roborock Saros Z70 for three months now and, in many ways, it's a great robot vacuum. It's the first robot vacuum with a mechanical arm meant to pick up objects and help you clean before it vacuums and mops. As someone who has three cats who leave toys everywhere, I found this incredibly appealing. I test and write about vacuums for a living and the Saros Z70 really is the first of its kind. CNET even chose it as one of our best of CES awards, giving me high hopes for its potential. I tested it at my cluttered apartment for three months and my colleague and lead lab tester Jared Hannah did the same at his home. The reality is, the Saros Z70 certainly has a lot of promise. It vacuums and mops like a champ, can cross thresholds and obstacles as effectively as the Dreame X50 Ultra, and has great mapping and pet recognition in my home. The OmniGrip arm picking up my wife's sock. Ajay Kumar/CNET However, we found its signature feature -- the robot arm -- is still not ready for showtime. Right now, the Saros Z70 is fairly limited in its ability to pick up very small objects like pet toys and children's toys. On occasion, it did recognize and attempt to pick up cat toys it found scattered around my home but it was never able to succeed. Often, it would deploy the arm then detect my cat or some other object nearby and freeze, blaring "Error 69," forcing me to manually reset the arm. At one point, it even tried to pick up my standing desk mat before realizing it wasn't a compatible object and giving up. I tested the Saros Z70 through June and July with Roborock's latest software updates that were intended to improve performance. Here are my impressions after three months of testing and whether it's worth its $2,600 price tag. Roborock My experience with the Saros Z70 We tested the Saros Z70 at CES 2025, at a live demo, at our New York City office space, at CNET's testing lab in Louisville, Kentucky, and most recently at my home for three months. My biggest takeaway from my real-world home test versus my other experiences with this robot vacuum is the various issues with the mechanical arm, its struggle to pick up objects, navigation problems and software bugs. It cleaned great as a regular robot vacuum and mop but I couldn't rely on the arm's sorting capabilities to reliably pick up shoes and socks, let alone cat toys. "Yeah, it seems like it's basically a beta version of the arm," said Jared, who was testing the Z70 at the same time. "They should do a bit more development before it's ready for the masses. I would get the error if the arm bumped something while moving an object. It sensed that there was too much resistance so I think it was doing what it was supposed to but it seems like that would end up being a common occurrence." The Saros Z70 got scratched up after trying to clean under the desk chair. Ajay Kumar/CNET Vacuuming and mopping: Strong performance, good sand/litter/hair pickup; mopping with clean/dirty tanks and hot air drying is effective. Strong performance, good sand/litter/hair pickup; mopping with clean/dirty tanks and hot air drying is effective. Special features: The OmniGrip arm is the standout special feature although currently unreliable. I found operational issues during my home testing. It struggled to pick up certain objects, especially pet toys, but I had much better luck with socks. Not only was the Saros Z70 able to identify my wife's purple socks and pick them up off the hardwood floor in the bedroom but it also began to transport them to my designated shoe-drop-off area that's set up in the foyer. The problem was that partway on this journey, it gave up somewhere between the hallway and the kitchen, dropping the sock, acting like the job was complete and continuing its regular cleaning cycle. The OmniGrip arm is the standout special feature although currently unreliable. I found operational issues during my home testing. It struggled to pick up certain objects, especially pet toys, but I had much better luck with socks. Not only was the Saros Z70 able to identify my wife's purple socks and pick them up off the hardwood floor in the bedroom but it also began to transport them to my designated shoe-drop-off area that's set up in the foyer. The problem was that partway on this journey, it gave up somewhere between the hallway and the kitchen, dropping the sock, acting like the job was complete and continuing its regular cleaning cycle. Mapping and navigation: It generated a 2D and 3D map of my space, automatically tagging the rooms (bedroom, bathroom, living room and kitchen), flooring type (hard surface and carpet) and identifying certain pieces of furniture (couch, TV stand and bed). It also marked notable thresholds in my space where extra effort would be required to cross. You can remap it or tell it to remap a specific room if you're not happy with the result but it did a great job for me the first time around. It generated a 2D and 3D map of my space, automatically tagging the rooms (bedroom, bathroom, living room and kitchen), flooring type (hard surface and carpet) and identifying certain pieces of furniture (couch, TV stand and bed). It also marked notable thresholds in my space where extra effort would be required to cross. You can remap it or tell it to remap a specific room if you're not happy with the result but it did a great job for me the first time around. Software and app: The app isn't the most straightforward tool to use, with a profusion of settings and nested menus that allow you to adjust just about everything the robot can do, from the dustbin emptying and mop washing frequency, to cleaning patterns, obstacle photos and enabling the robotic arm. Just about everything the robot can do is customizable but I wish all these features and settings were a bit more front and center rather than tucked away in the lengthy settings menu. The app isn't the most straightforward tool to use, with a profusion of settings and nested menus that allow you to adjust just about everything the robot can do, from the dustbin emptying and mop washing frequency, to cleaning patterns, obstacle photos and enabling the robotic arm. Just about everything the robot can do is customizable but I wish all these features and settings were a bit more front and center rather than tucked away in the lengthy settings menu. Design and aesthetics: Low-profile design, good for fitting under furniture, but can lead to scratches when it tries to wedge itself under things like my wife's office chair. Low-profile design, good for fitting under furniture, but can lead to scratches when it tries to wedge itself under things like my wife's office chair. Ease of use: Setup had connectivity issues; once set up, basic functions are easy, but advanced settings are complex. Roborock The specs Price: $2,599 $2,599 Object weight limit (OmniGrip Arm): 300 grams rated load; approximately 700 grams actual load capacity during our testing. 300 grams rated load; approximately 700 grams actual load capacity during our testing. Threshold crossing height: Capable of crossing thresholds up to approximately 2 inches. Capable of crossing thresholds up to approximately 2 inches. Cleaning time: (Ajay's apartment, 850 sq ft): 1.5 to 2 hours for a full vacuum and mop session. (Ajay's apartment, 850 sq ft): 1.5 to 2 hours for a full vacuum and mop session. Cleaning time: (Jared's house, 1,200 square feet) : 2.5 to 3 hours for a full vacuum and mop session on the main level. (Jared's house, 1,200 square feet) 2.5 to 3 hours for a full vacuum and mop session on the main level. Mapping time (Ajay's apartment, 850 sq ft): 16 to 18 minutes for initial mapping. (Ajay's apartment, 850 sq ft): 16 to 18 minutes for initial mapping. Sand pickup performance on hardwood: 82.46%, this is excellent hardwood performance, falling just shy of our best overall, the Ecovacs Deeboth T30S Combo (85%). 82.46%, this is excellent hardwood performance, falling just shy of our best overall, the Ecovacs Deeboth T30S Combo (85%). Sand pickup performance on carpet (mid-pile): 51.77%, this surpasses the Dreame X50 Ultra (44% mid-pile) 51.77%, this surpasses the Dreame X50 Ultra (44% mid-pile) Sand pickup performance on carpet (low-pile): 51.06%, which comes close to matching the Shark Power Detect NeverTouch Pro (54.5%). 51.06%, which comes close to matching the Shark Power Detect NeverTouch Pro (54.5%). Voice command compatibility: The app can be connected to Amazon Alexa for voice commands. The app can be connected to Amazon Alexa for voice commands. Customizable: The app allows for a range of customization, including dustbin emptying frequency, mop washing frequency, cleaning patterns, obstacle photos and enabling/disabling the robotic arm. The app allows for a range of customization, including dustbin emptying frequency, mop washing frequency, cleaning patterns, obstacle photos and enabling/disabling the robotic arm. Mapping features: The app displays the 2D and 3D maps generated by the robot, allows for room tagging, furniture identification and setting "no-go" zones. It also identifies pet areas and allows for "Pet Area Cleaning." The app displays the 2D and 3D maps generated by the robot, allows for room tagging, furniture identification and setting "no-go" zones. It also identifies pet areas and allows for "Pet Area Cleaning." Wi-Fi connectivity: The robot needs to connect to Wi-Fi for setup and app control. The robot needs to connect to Wi-Fi for setup and app control. Dual spinning mopping pads: The robot uses a pair of dual spinning mopping pads that attach and detach from the docking station. The robot uses a pair of dual spinning mopping pads that attach and detach from the docking station. Clean and dirty water system: Uses clean water for mopping and collects dirty water in a separate tank, ensuring cleaner mopping. Uses clean water for mopping and collects dirty water in a separate tank, ensuring cleaner mopping. Mopping pad cleaning and drying: Mopping pads are washed and hot-air dried at the base station between cleaning sessions. Mopping pads are washed and hot-air dried at the base station between cleaning sessions. Mopping modes: The "Vacuum+Mop" mode vacuums first and then mops the entire space. There are options for Deep cleaning, Pet Area and an AI-powered SmartPlan. Saros Z70 docked Ajay Kumar/CNET CNET's buying advice While the Roborock Saros Z70 excels at vacuuming, mopping and navigation, the core feature -- the OmniGrip arm -- is not yet reliable enough for me to recommend it if you're only looking to make heavy use of the OmniGrip arm. Recent software updates in June and July have made some improvements, including improved grab point optimization, smarter sorting optimization and new pickup categories like sports shoes but the consistency still leaves a lot to be desired. It's hard to ignore the $2,599 price tag. I argue that for the price, you could purchase two or three excellent standard robot vacuums and mops, like the Yeedi M12 Pro Plus, or a comparable high performing model such as the Dreame X50 Ultra which has the ability to cross thresholds -- both on CNET's best robot vacuums best list. On the even lower price end, iRobot's new Roomba 205 DustCompactor Combo is our best new robot vacuum for less than $500, with solid performance and a unique mechanical paddle that lets you store 60 days of dust in the dustbin. If all you care about is strong vacuuming, mopping and obstacle avoidance: The Ecovacs Deebot T30S Combo is currently CNET's "best overall robot vacuum" and Roborock's own Saros 10R offers similar or better vacuum and mopping performance with more space for the dustbin and water tank because it doesn't need to fit a mechanical arm in its body. Roborock Without a fully operational and reliable OmniGrip arm, I can't justify its value for the cost at this time. "The arm is a neat feature to watch but I personally don't find it very useful," said Jared. "I feel like if the robot is smart enough to avoid obstacles such as shoes and socks, I don't need it to try and pick them up and move them. As long as the robot doesn't get stuck on anything during its cleaning cycle, that's good enough for me." If the OmniGrip arm is your primary reason for considering the Saros Z70, I recommend you continue to wait for future software updates or wait a few years until Roborock and other companies enter the market with newer generations of the technology.

How AI Is Driving Innovation In The Automotive Industry
How AI Is Driving Innovation In The Automotive Industry

Forbes

time3 days ago

  • Forbes

How AI Is Driving Innovation In The Automotive Industry

Rajnish Nath, President of Manufacturing, Automotive, Aerospace & Defense and Life Sciences at Capgemini Americas. For years, pop culture and various media have envisioned our cars of the future. From the DeLorean of the "Back to the Future" franchise to The Jetsons' aerocar, these portrayals have offered a window into how revolutionary automotive technology could reshape the way we travel, live and interact with vehicles in the future. Today, flying cars are being designed by a variety of start-ups, helping to push the boundaries of mobility. Each day, we move closer to fully autonomous, incredibly intelligent and connected vehicles. AI is fundamentally changing the automotive sector, from product design, supply chain and manufacturing, safety and logistics, all the way down to the in-cabin experience. It seems like there is virtually no limit to AI and other transformative technologies' ability to drive innovation. AI's Role In The Supply Chain Although the current geopolitical uncertainties, supply chain disruptions and economic pressures have the automotive industry scrambling to adapt quickly and efficiently, I believe there is an exciting opportunity for organizations to rethink their factories and supply chains of the future. In today's political headwinds, every organization must ensure the longevity and resilience of its business in the market, with 62% of organizations acknowledging the importance of establishing domestic manufacturing and supply chain infrastructure to strengthen local capacity. Arguably, one of AI's most important roles is its potential to revolutionize supply chain operations. AI-powered technologies can provide companies with the power to better map their supply chains, detect changes in supply and demand, and recommend proactive measures to mitigate disruptions earlier. For example, AI technology is already helping automotive OEMs adjust their strategies to improve part supply by understanding multiple market signals to continuously monitor the risk of parts shortages to keep plant lines running. Early guidance can enable alternate sourcing and reduce or eliminate reactive approaches. Hand in hand with other advanced technologies such as agentic AI, machine learning (ML) and large language models (LLMs), AI-powered technologies are helping companies mitigate supply chain risk to continue manufacturing despite market shifts. Additionally, the rise of hybrids and electric vehicles (EVs) has added complexity to the automotive supply chain, accounting for a record 20% of US vehicle sales in 2024. Beyond working together to improve planning, forecasting and decision making, these tools will be crucial in supporting the manufacturing of complex vehicles and their software-rich components to keep up with the ever-changing customer demand. While AI will be key in building proactive resilience and risk management in the automotive supply chain and manufacturing, we're also seeing how AI-powered innovations are powering the vehicles of our future. AI-Powered Innovations In Modern Vehicles Nearly every major automaker around the world utilizes advanced driver assistance systems (ADAS) in some form. With the global ADAS market projected to grow from $72.7 billion in 2025 to $260.5 billion by 2035, it's clear that industry leaders see this technology as essential for enhancing vehicle safety, boosting sales and driving new revenue streams through connected services. AI plays a central role in the evolution of ADAS. Its ability to process vast amounts of data, interpret complex environments and support real-time decision-making brings us closer to fully autonomous vehicles. Currently, 26% of automotive organizations are using generative AI to accelerate autonomous vehicle development. Looking ahead, we can expect to see even greater use of neural networks, proprietary algorithms and agent-based AI in engineering software-defined vehicles. These technologies aim to achieve faster time to market, greater reliability and safety, and improved customer satisfaction—all while keeping human engineers at the wheel. AI is also transforming the in-cabin experience by making it safer and more personalized. Features like voice-activated controls as well as tailored recommendations for food, fuel and eco-friendly routes are already enhancing daily driving. Now, we're seeing next-level personalization with innovations such as interior lighting designed to reduce motion sickness and holographic dashboards, turning cars into immersive, personalized living spaces. Beyond the driving experience, AI is also streamlining post-sale operations. It's helping manufacturers and suppliers close the warranty feedback loop more quickly by identifying product issues through both customer reports and vehicle telematics. This accelerates root cause analysis and improves product design. In service environments, AI tools are helping technicians improve repair accuracy by analyzing shared images, videos and sounds of vehicle issues. Soon, AI agents will be capable of providing real-time guidance on warranty coverage simply by reviewing a media file and responding to technician queries, driving faster, more accurate and more transparent vehicle servicing. The Road Ahead Over the next decade, AI is poised to play a transformative role in the automotive industry, reshaping vehicle development, enhancing driving experiences and influencing urban mobility. As software becomes increasingly central to vehicle functionality, the shift toward software-defined vehicles will accelerate, gradually moving away from traditional engineering paradigms. While fully autonomous vehicles may not yet dominate the roads, we can expect more widespread integration of ADAS, contributing to incremental safety and performance improvements. In-car experiences are also likely to become more adaptive, with AI personalizing settings and preferences to individual drivers. This period of technological transition brings both uncertainty and opportunity. AI holds promise for strengthening supply chain resilience, offering predictive capabilities that could help automakers navigate disruptions with greater agility. At the same time, its role in connecting vehicles to smart city infrastructure opens new possibilities for traffic optimization and urban mobility services, such as autonomous ride-hailing and shared transport. In today's automotive landscape, AI is not just a tool; it's becoming a foundational enabler of innovation. From engineering and manufacturing to the user experience, AI is helping to redefine what vehicles can be, setting new benchmarks for safety, customization and customer expectations. Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?

DOWNLOAD THE APP

Get Started Now: Download the App

Ready to dive into a world of global content with local flavor? Download Daily8 app today from your preferred app store and start exploring.
app-storeplay-store