智通财经网·2025-08-17 23:48Core Insights - The report from CITIC Securities highlights that humanoid robots (non-completely humanoid) need to possess multimodal perception and end-to-end large model capabilities to achieve logistics sorting operations [1][2] - Current hardware for humanoid robots has reached the commercialization threshold in logistics scenarios, but domestic embodiment models still require improvement for practical application [2][3] - The economic analysis indicates that the cost-benefit ratio of humanoid robots is now on par with sorting workers, assuming a two-year cost recovery period [3] Economic Analysis - In logistics sorting, a single worker working two shifts (8 hours each) earns an annual salary of 100,000 yuan, while the cost of a humanoid robot is 400,000 yuan, working 20 hours a day with an efficiency of about 80% compared to workers [3] - The cost-benefit ratio of humanoid robots is expected to align with that of human workers as hardware costs decrease and efficiency improves [3] Technological Development - The sorting actions performed by humanoid robots include package detection, grabbing and flipping, barcode recognition, path planning, and delivery positioning, requiring advanced multimodal perception and autonomous decision-making capabilities [2] - The Figure 02 robot, utilizing Helix neural network, processes a package in an average time of 4.05 seconds, comparable to skilled sorting workers who take 3-5 seconds per item [2] - The domestic robot, Zhiyuan Jingling G1, still lags behind skilled workers in sorting speed and quality [2] Future Outlook - In the second half of this year, humanoid robots are expected to transition from demo scenarios to customer trials, with potential for explosive growth in the following year after customer validation [3] - The logistics sector abroad, where labor costs are higher, is anticipated to be the first to adopt these technologies [3] - Companies with strong secondary development capabilities and those leading in commercialization progress are recommended for attention [3]