As dawn breaks, 45 intelligent street sweepers automatically depart from the sanitation depot in the Beijing Economic-Technological Development Area (Beijing E-Town), slowly making their way to preset routes to begin their daily work. According to the district's Urban Operations Bureau, more intelligent cleaning machines are expected to come online before the end of the year, covering over 200 roads across the entire 60-square-kilometer area. The shift from manual brooms to machine-driven cleaning is rewriting how urban sanitation is carried out, turning the vision of "technology improving life" into everyday reality.
On Hongda North Road, one of these small intelligent sweepers working along a pedestrian walkway has drawn widespread curiosity. Standing just half a person's height and less than a meter wide, it resembles a round, sturdy "iron buddy." During operation, its front side brushes extend outward, spinning as they gather various small bits of debris toward the undercarriage, where suction nozzles "swallow" the waste into a dust collection bin. At the same time, it sprays a fine mist of water to suppress dust from rising into the air.
The "iron buddy" performs its duties thanks to new equipment and millimeter-level algorithms: cameras around its body give it a 360-degree view of the road surface, allowing it to hug the right curb while cleaning; a lidar sensor on its roof spins silently, enabling it to deftly avoid obstacles; and if a person or pet suddenly appears, it immediately slows down and brakes, quietly yielding the way.
"It may move at a leisurely pace, but it cleans quite thoroughly. We watched as the fallen leaves on the road were simply sucked away without kicking up any dust," remarked an elderly couple walking behind the machine after buying groceries. These autonomous intelligent sweepers first appeared in August last year. After verification of cleaning performance and safety obstacle avoidance, the E-Town district partnered with three companies—WeRide, Cowell Technology, and Feiba Technology—to formally deploy 45 of these "iron buddies" in a pilot program. All vehicles have obtained road testing permits for intelligent connected sweeping vehicles, with 10 medium-sized 6-ton units capable of working around the clock on motor vehicle lanes and non-motorized lanes. The other units, weighing less than 1 ton, are currently used only during daytime for pedestrian walkway operations.
Sheng Xutao, general manager of Beijing E-Town Zhijie Environmental Sanitation Technology Co., Ltd., explained that aside from extreme weather conditions and other exceptional circumstances, these intelligent sweepers now hit the road every day according to operational plans and dispatch schedules, covering 78 roads within the 10-square-kilometer core zone of E-Town. To evaluate their actual performance, the district recently conducted a "man vs. machine" comparison using the manually cleaned Jingyuan Street as a baseline. A professional testing company acted as the "examiner," randomly measuring residual dust levels on motor vehicle lanes of the pilot roads each week. The results showed that the "iron buddies" excelled at dust collection, with residual dust levels basically meeting Beijing's road cleanliness standards after every operation, while visible litter on pilot roads remained stable at 2 to 3 items per 100 square meters. The comparison on pedestrian walkways was even more compelling: manual sweeping achieved a dust removal rate of less than 20%, while the intelligent equipment reached 60 to 70 percent.
"It's like the difference between using a broom and a vacuum cleaner when we sweep at home," Sheng Xutao offered a vivid analogy. On the same walkway surface, the intelligent equipment can collect "three to five extra loads" of dust compared to manual labor. With less dust and debris, overall street cleanliness improves, directly enhancing the commuting and living experience for residents.
Pedestrian walkways were once considered a "forbidden zone" for traditional street sweepers due to their narrow width, high curbs, and obstructions like shared bikes, electric scooters, and tree pits—large vehicles simply couldn't navigate them. Now, the "iron buddy" has changed that, though its obstacle avoidance can leave behind missed spots. "We identified this issue during the pilot and have developed an optimal human-machine collaboration path," explained Cui Huasheng, deputy director of the E-Town Urban Operations Bureau. Intelligent equipment handles large-scale routine cleaning, while manual workers cover corners, narrow areas, unexpected situations, and fine-detail work. This division has significantly reduced the repetitive labor burden on sanitation workers.
"In the past, sweeping walkways meant wielding oversized bamboo brooms that weighed over ten jin (about 5 kilograms) each. After a full day, my arms would be sore," said sanitation worker Wang, who now partners with the "iron buddy." He recalled that windy days were the worst—the broom would stir up a cloud of dust that made people cough, and many passersby would cover their mouths and hurry past, or simply cross to the other side of the street. "Now it's much easier," he said, holding up a small broom no bigger than a household one. "They patrol their assigned routes two or three times a day. We just need to fill in the gaps, sweeping corners, brick crevices, and around tree pits. The old giant broom isn't needed anymore."
The "iron buddy" not only works hard but is also tireless—as long as it has power, it can be dispatched 24/7. During nighttime and high-temperature hours, it takes the lead, greatly reducing the physical strain on sanitation workers. Traditional cleaning vehicles and personnel are now dedicated to water spraying, dust suppression, and emergency operations during leaf-fall season or strong winds, which extends cleaning coverage hours and reduces safety risks. After more than two months of testing, the intelligent equipment has shown steadily improving cleaning quality on pilot roads with clear advantages in dust control, capable of replacing one to two manual crews for large-area sweeping. The "human-machine collaboration" model leverages the dual benefits of large-scale, standardized machine cleaning and precise manual touch-ups, improving both quality and efficiency. It provides a replicable operational paradigm for similar cities and drives the transformation of sanitation work from labor-intensive to intelligent, compact operations.
After more than a year of trials, the "iron buddy" has made significant progress: operations have shifted from "manual intervention-first" to "system automatic dispatch and equipment autonomy," with increasing automation and fewer human touchpoints, while standardization and stability have improved notably. Currently, aside from three tasks—water refilling, charging, and waste discharge—that are limited by depot infrastructure, the "iron buddy" has achieved a closed-loop intelligent workflow covering automatic dispatch and retrieval, standardized sweeping, and automatic waste collection.
"Intelligent cleaning is just the entry point," Cui Huasheng noted. The E-Town district is using this as a breakthrough to explore broader intelligent approaches to fine-grained urban governance. At present, the district is conducting feasibility studies on smart retrofitting of cleaning supply stations and guiding the development of an integrated "Smart Sanitation Steward" management platform, which will enable unified dispatching, unified standards, and unified assessment across cross-company fleets, ensuring all operations are measured by "one ruler" throughout the entire area. According to the plan, after the core zone pilot concludes in September, intelligent cleaning equipment will begin expanding district-wide across the full 60 square kilometers toward the end of this year, transitioning into normal operations—a rare case of "district-wide implementation and large-scale operation" nationwide. In preparation for the upcoming large-scale deployment, E-Town is now preparing for full-caliber comparative testing, cost estimation, and standard formulation across multiple devices and scenarios. When fully rolled out, more "iron buddies" will be deployed with differentiated configurations based on road class, achieving full-scenario coverage from main arteries to branch roads and from daytime to nighttime operations.