While logistics companies like Amazon and GXO Logistics are increasingly using humanoid robots in warehouses to assist with material handling, FedEx CEO Raj Subramaniam has a different perspective.Different opinionsHe believes that ordinary robots that simply mimic human appearances are simply not capable of handling the complex working environment inside FedEx warehouses.
His ideal type is a "super humanoid robot" with more joints, or even "several elbows".
Why isn't being "human-like" enough?
In an interview with The New York Times, Raj Subramaniam stated frankly that humanoid robots are not yet capable of performing the job for FedEx.
"Loading and unloading cargo on trucks is a very difficult problem for robots," Raj Subramaniam explained, "because packages come in all sizes, shapes, and weights."
While the human body is flexible, the limitations of hands and a single elbow joint become a bottleneck when faced with a freight car filled with cluttered packages. Therefore, FedEx is not seeking robots that simply mimic humans, but rather "superhuman robots." "Perhaps they need to have multiple elbows and more degrees of freedom," explains Raj Subramaniam.
However, he also admitted that FedEx's application of such robots is still in the pilot stage and is still some distance away from being officially launched.
AI predicts 1700 million packages per day
In addition to hardware robots, FedEx is also fully investing in AI software applications. Raj Subramaniam stated that the company is using data from the 1700 million packages shipped daily to train AI models to more accurately predict delivery times and improve customer experience.
Industry perspective: Humanoid vs. Specialized
FedEx's cautious approach contrasts sharply with its competitors. Amazon currently has over 75 robots assisting with picking and moving, and is actively testing humanoid robots, including Agility Robotics' Digit. Logistics company GXO Logistics has also stated that it is actively deploying humanoid robots, and Morgan Stanley even predicts that by 2030, robots could save Amazon $100 billion annually in costs.
However, FedEx's view is not unique. In 2023, the chief technology officer of Boston Dynamics also stated that for certain warehousing tasks, a human body may not be the most efficient solution.
Qualcomm also holds a similar view.I do not believe that robots must be designed in a humanoid form.Anshuman Saxena, Vice President and General Manager of Automation and Robotics Solutions at Qualcomm, believes that humanoid robots are indeed the ultimate form for general applications. After all, mimicking the human form and the movements of two hands and two fingers will enable them to be applied in many industrial applications or quickly transferred to different market demands, unlike robots designed for warehousing systems, which have greater limitations in deployment and application in other fields.
Analysis of viewpoints
The views of the FedEx CEO point out a fundamental debate in the current robotics industry: "General Purpose" versus "Specialized" considerations.
The current craze for humanoid robots like the Tesla Optimus and Figure 01 is based on the logic that "the world was designed for humans, so robots that look like humans are the most universal." However, in highly demanding and efficient environments like logistics and warehousing, human physiology is actually full of limitations (for example, we cannot bend our arms backward or reach into deep crevices).
The "multiple elbows" mentioned by the FedEx CEO may sound like something out of a science fiction horror movie, but they make perfect sense from an engineering perspective. If a robotic arm can bend as flexibly as a snake (redundant degrees of freedom), it can avoid obstacles and grab packages from corners without moving its body. This also shows that for the efficiency-driven logistics industry, they don't care whether the robot looks like a human; they only care whether it can be stronger, more unconventional, and better at moving things than a human.
