Friday, 28 November 2025
25.3 C
Singapore
16.7 C
Thailand
23.4 C
Indonesia
27.8 C
Philippines

Humanoid robot sets new long-distance walking record in China

A humanoid robot from AgiBot has broken a walking record in China, completing 106.3 km in three days while raising key technical questions.

A humanoid robot developed by the Chinese robotics company AgiBot has set a new world record, completing the longest walk ever by a machine of its kind. The full-size robot, known as A2, covered 106.3 kilometres over the course of three days, finishing its journey along Shanghai’s well-known Bund waterfront promenade.

The walk took the robot from Suzhou to Shanghai, passing through busy urban areas, travelling alongside national highways, and crossing several bridges. Video released by AgiBot shows the robot moving steadily and maintaining balance across changing environments. The achievement is notable not only for the distance covered but also for the robot’s ability to operate continuously without rest, made possible by AgiBot’s rapid hot-swap battery system.

Wang Chuang, senior vice president of AgiBot, commented on the event, saying: “Walking from Suzhou to Shanghai is a tough task even for many humans, but the robot achieved it.” He added: “It proves the maturity of the robot’s hardware, cerebellar balance algorithms, and endurance, laying a solid foundation for large-scale commercial deployment.”

Record-breaking walk raises technical questions

Although the accomplishment has attracted global attention, several questions remain about how the challenge unfolded. AgiBot has released limited details about the process, leaving observers keen to learn more about the operation behind the scenes.

One key area of interest is the speed and efficiency of the hot-swap battery mechanism. AgiBot has not confirmed how many battery changes occurred or how long each swap took. Seeing the procedure in action would offer useful insight into the robot’s readiness for real-world deployment, particularly in scenarios that require continuous activity.

Another technical question concerns the navigation process. While the company is expected to have mapped the route in advance, it is unclear whether any real-time adjustments were needed. Understanding how the robot responded to unpredictable environmental factors could provide a clearer picture of its ability to adapt autonomously to dynamic surroundings.

Monitoring, maintenance, and on-route performance

It also remains unknown how closely the robot was monitored during the walk. It is unlikely that the A2 completed the journey entirely unaided, but AgiBot has not disclosed the size of the support team or its exact role. Clarifying whether the robot travelled with a full crew, a small technical team, or minimal human oversight would shed light on how much supervision the technology still requires.

AgiBot has similarly not addressed whether any maintenance was carried out during the three-day trek. With complex mechanical systems operating over extended distances, maintenance needs can reveal important details about the robot’s durability. The company has not said whether the A2 experienced any breakdowns or if engineers learned anything new about its long-term resilience throughout the journey.

Challenges faced on the route

There is also interest in whether the robot encountered any specific challenges. With obstacle-avoidance sensors built into its design, the A2 is equipped to handle everyday environments, but the real world is rarely predictable. As the robot walked among pedestrians, vehicles, and varying road conditions, it was likely to encounter new or unexpected situations. AgiBot has yet to provide information on how the robot responded to such scenarios or whether any issues occurred.

Despite these unanswered questions, the record-setting walk marks an important milestone for the robotics industry. AgiBot joins a growing field of companies racing to design humanoid robots capable of assisting in workplaces and, ultimately, in homes. While rapid progress is being made, experts note that large-scale production and meaningful integration into the workforce remain several steps away.

Hot this week

Valve offers strongest hint yet on expected Steam Machine pricing

Valve hints that the Steam Machine may be priced close to a similarly powerful DIY PC, but external factors keep final costs uncertain.

Global mobile gaming ads surge in 2025 as AI and interactivity reshape engagement

Mobile gaming ads grew strongly in 2025 as AI-driven optimisation and interactive formats reshaped global user acquisition strategies.

Meta explores an AI briefing tool aimed at Facebook users

Meta is developing Project Luna, an AI tool designed to deliver personalised morning Facebook briefings to users.

Andika Rama returns to claim TGR Asia Esports GT Championship 2025 title

Indonesia’s Andika Rama wins the TGR Asia Esports GT Championship 2025 as his team seals both individual and country titles.

Belkin Zootopia accessories you need before Zootopia 2 arrives

Belkin’s latest Zootopia collection brings fun designs and practical features to power banks, cables, cases and straps for everyday use.

Singapore orders Apple and Google to stop spoofed government identities on messaging apps

Singapore orders Apple and Google to block spoofed government identities on messaging apps to curb rising impersonation scams.

Nintendo acquires Bandai Namco Studios Singapore

Nintendo acquires Bandai Namco Studios Singapore to boost game development and expand its subsidiary network.

Google DeepMind opens new AI research lab in Singapore to strengthen regional language capabilities

Google DeepMind opens a new AI lab in Singapore to boost regional language understanding, research partnerships, and real-world innovation.

AMD powers Zyphra’s large-scale AI training milestone

Zyphra trains its ZAYA1 foundation model entirely on AMD hardware, marking a major step for large-scale AI development.

Related Articles