Why are there so many robots at a present targeted on telephones? That is the query I requested myself as I roamed the halls of Cell World Congress final month, looking out for essentially the most thrilling know-how that can outline the subsequent few years.
The primary and most blatant reply is that robots draw crowds. A dancing humanoid is a straightforward technique to appeal to folks to your sales space. However to see the robots at this 12 months’s MWC purely as a publicity stunt can be to disregard the larger dialog taking place round robots and connectivity.
Already in 2026, we have seen main leaps ahead in robotics, with corporations together with Boston Dynamics and phone-maker Honor displaying off humanoid robots designed for business and residential environments. However there’s yet one more degree to unlock, and it depends on 6G — the next-generation community know-how set to succeed 5G in 2030 and past.
On the floor, 6G and robotics might sound distinctly unrelated — past being applied sciences of a future that we’re not residing in fairly but. However on this future, 6G will open new doorways for humanoid robots that’ll rework them from clunky, standalone mechanical collectible figurines into environment friendly fleets, the place people will type a part of an all-sensing, always-learning ecosystem.
This may occur first in business, then in hospitality and care environments, earlier than doubtlessly touchdown in our houses. It is an thrilling prospect, however because the consultants I spoke to at MWC final month cautioned, there will be some massive leaps in know-how required earlier than they, and we, are prepared for that.
The ability of 6G
To know how 6G will unlock new potentialities for robots, let’s begin with the particular capabilities the community know-how may have.
The primary is that 6G will act as a sensor community, with sensors embedded into each the robots and their environments, Qualcomm’s govt vp of Robotics Nakul Duggal instructed me.
This enables the 6G radio to behave like radar — continually scanning and mapping its environment in actual time to detect obstacles. Think about a robotic trying to navigate a crowded surroundings: The 6G community ought to rapidly and cheaply assist create a form of digital map for it to take action safely.
Second, there’s the pure pace at which 6G will talk huge reams of information. The 5G networks we at the moment use aren’t essentially constructed to deal with AI requests, however the 6G networks can be, offering a constant, low-latency, comparatively low-power technique to course of intelligence and ship that intelligence to robots, in keeping with Frank Lengthy, affiliate director of clever companies at deep tech analysis agency Cambridge Consultants.
Personal 5G networks mixed with edge AI (counting on gadgets for computing, not simply the cloud) can fill the hole for now, however public networks, not a lot. Against this, Lengthy stated, “with 6G you possibly can just about have that high quality of service assure.”
Cambridge Consultants introduced a demo of an autonomous humanoid robotic to MWC that may decide up and place down a field based mostly on the place it sees you pointing. The gesture recognition, plus the flexibility to react in actual time, whereas various its grip to choose up one thing that is likely to be on an angle, requires an unlimited quantity of compute energy. (The demo was powered by a personal 5G community.)
The robotic was in a position to decide up this field and place it on a spot I pointed to.
Whether or not robots are linked to the cloud, or to one another in a peer-to-peer fleet, the community might want to deal with their intelligence calls for at pace. For robots to be continually speaking to the infrastructure round them — and to one another — a powerful, dependable uplink can be required, defined Anshuman Saxena, basic supervisor of robotics at chipmaker Qualcomm.
He gave the instance of two robots working in a retail surroundings the place one is unloading soda cans from a truck, and one other is restocking cabinets. They’re going to must align on find out how to learn the house round them to finish every job, together with understanding what number of cans will want inserting, and once they’ll be prepared.
“The one means is that this robotic, whereas shelving, goes to the again door entry of the truck that’s getting unloaded and sees what is accessible,” stated Saxena. “Or the robotic that is unloading is speaking the larger image to each different robotic, in order that we now have a view of the place the issues are positioned, in order that they’ll plan.”
That is what’s often called long-horizon planning, the place a robotic is not simply specializing in the fast job however occupied with how that job matches right into a broader context over an extended timeframe inside a dynamic and unstructured surroundings. In different phrases, it is performing the form of ongoing psychological multitasking that people do each day, reacting at pace to what is going on on round us, whereas additionally contemplating what’s subsequent. Within the Cambridge Guide demo, the robotic was able to pondering 16 steps forward.
In the meantime, lightning-fast 6G will assist robots make split-second selections, based mostly on suggestions not simply from their very own sensor-packed our bodies, however from different robots and tech within the surroundings. “The retail shops have cameras,” stated Saxena. “It is not a robotic, however it may be the eyes of the robotic.”
For robots, on daily basis can be a college day
In your individual dwelling, you may need solely a single humanoid robotic. However that will not be as totally different from the retail situation as it’s possible you’ll assume.
That is as a result of most of the gadgets you personal, together with your cellphone and safety cameras, can already talk with one another, and the robotic can be simply one other one within the combine. Or perhaps you will have one humanoid and a bunch of smaller robots designed for particular duties.
“There’s a fleet facet within the merchandise that we use,” Duggal stated. “You do not really feel that, however that’s precisely how the product is working.”
Understand that your cellphone is each a bodily object itself and all of its software program and knowledge are managed elsewhere. The cellphone additionally supplies suggestions to refine that software program, as will the 6G-equipped robots.
“So a robotic goes to be performing a sure bodily job, and whereas it might carry out it in your house, if it is also performing the identical job in lots of different houses, there’s this facet of studying and deployment,” Duggal stated.
This steady studying is probably one of many greatest challenges that 6G is predicted to assist clear up in robotics. Robots and AI will want large quantities of real-world knowledge that in the present day’s networks cannot sustain with, even for mundane duties.
For instance: choosing up and serving you a cup of espresso, which entails dexterity and stability, with the added aspect of warmth. A robotic arm may not care in regards to the temperature. “However whether it is scorching, how would we react?” stated Saxena. “We’d simply rapidly go away it, which is a really quick response time.”
The pace of 6G networks can be important. By the point a robotic arrives in our houses, we are going to need to know that it should not hand us a scalding-hot drink and find out how to defend itself from injury.
A lot of this studying may need taken place in resorts or eating places, the place in a single day, robots load and unload dishwashers and reset the kitchen. The robotic will deliver that coaching into your private home, the place it will nonetheless must additional study your distinctive structure and routine. This may probably be a time-consuming course of.
Qualcomm is working with a number of robotics corporations, together with Neura Robotics, which develops robots for each industrial and residential use.
“It should be extremely difficult,” stated Lengthy. “Put it this manner, members of my fast household nonetheless wrestle with opening the newborn gate in my stairs, even after in depth coaching. So a robotic, I believe, is likely to be a number of years away from opening that child gate.”
Readying robots for 6G… and our houses
However 6G isn’t anticipated to roll out extensively till not less than 2030. What are the robots that corporations are already constructing and deploying to do till then?
They’re making the leaps and bounds they’ll with the networks of in the present day. “So you are not ready for 6G,” Saxena stated, “however when the connectivity comes alongside, you might be speaking about experiences which might be means past what robotics can do [today].”
Whereas the confluence of robotics and 6G will certainly unlock some hitherto unseen next-level robotics, there’s lots that robots can study within the meantime — significantly in relation to bettering dexterity — to prime them to benefit from higher connectivity. That is very true if we’re ever to think about inviting humanoids into our houses, an concept that feels, not less than for now, like one thing price delaying till not less than the 6G-enabled 2030s — if not past.
