Robots Handle the First 80%: RECONOVA Calculates the Automation Economics for Airports

When embodied intelligence enters real-world scenarios, the primary challenge is an economic calculation.

In the laboratory, robots can continuously challenge themselves with more complex movements; however, in real production environments, companies are more concerned with whether the robots can keep up with the production rhythm and how much existing infrastructure needs to be retrofitted.

This determines that robotic commercialization may not start with “100% automation,” but rather with tasks that are the most standardized, repetitive, and easiest to justify economically.

Airports, for instance, are particularly suitable for validating this logic. The volume of standard luggage is large, and task repetition is high, making the value of automation relatively clear. However, significantly increasing the complexity and cost of the entire robotic system to handle the remaining small amount of soft bags and irregularly shaped luggage may not be cost-effective.

At the 2026 World Robot Conference, which opened on August 19, RECONOVA fully showcased its airport luggage transfer robotic solution to the market.

The Xiaoyi luggage transfer robot primarily handles standard luggage, autonomously completing identification, grasping, transportation, and loading. Autonomous Mobile Robots (AMRs) are responsible for pallet transfer. For luggage with more complex forms, such as backpacks and soft bags, wheeled dual-arm robots are employed.

RECONOVA does not expect robots to solve every problem. Currently, in the real flight support environment at airports in East China, approximately 80% of standard luggage is handled by Xiaoyi robots, while the remaining 20%—consisting of soft bags, irregular shapes, damaged items, and other anomalies—is still handled through human-robot collaboration.

RECONOVA disclosed that during Proof of Concept (POC) testing in real flight support environments, the Xiaoyi robot’s cycle time for processing a single piece of luggage was under 18 seconds. The maximum loading capacity per vehicle was 39 pieces, approaching the 40-piece loading level of skilled human workers under similar conditions, with a loading accuracy rate of 99.9%.

Looking ahead, as models and other capabilities mature, RECONOVA plans for specialized robots to handle 80% of standard, high-frequency tasks. Robots with greater generalization capabilities will gradually take on 10% of complex, flexible tasks, while humans will remain the fallback for the final 10% of extreme long-tail tasks.

Behind this ratio lies a more realistic path for robotic commercialization: assign to robots only those tasks where the technology is mature and the economics make sense.

For airports, this is more practical than focusing on how many “showy” maneuvers a robot can perform. Airports already have conveyor belts, tugs, and operational processes that have been in place for decades; it is impossible to overhaul the entire system just to deploy robots.

By emphasizing “adapting robots to the scenario,” RECONOVA is essentially reducing the cost for customers to deploy automation.

This approach may well become a relatively feasible solution for deploying robots in the real world in the future.