Korea’s logistics giants are not waiting for foreign humanoids to arrive fully formed. Lotte is folding a national R&D project into its own fulfilment centres—and expanding a prior Jincheon pilot into chained multi-step work.
On 7 October 2026, Chosun Biz reported that Lotte Global Logistics was selected for the “2026 robot industry technology development project” led by the Korea Planning & Evaluation Institute of Industrial Technology (KEIT) under the Ministry of Trade, Industry and Resources (MOTIE). The project period is about five years, from this year to March 2031. [1]
What the national project is for
The stated goal, per Chosun Biz, is to develop a domestic physical AI model usable on logistics sites—reducing reliance on foreign-made robots and AI models and establishing a foundation for self-reliance in AI and robot technology in Korea’s logistics industry. [1]
Lotte Global Logistics will carry out technology development with Yujin Robot, CISO, and the Korea Electronics Technology Institute (KETI), which are also described as part of the “manufacturing AX (M.AX) alliance.” Lotte Himart joins as a demand-side company. The plan is to combine each organisation’s robot hardware manufacturing technology, AI algorithms, and on-site piloting capabilities. [1]
From one apparel pilot to a continuous sequence
Lotte Global Logistics last year piloted a process from apparel picking to polybag packing using an AI bipedal humanoid at the Jincheon fulfilment centre. In this project, the company says it will expand autonomy so the humanoid performs a continuous sequence—order confirmation, product picking, quantity verification, placement, transport, barcode confirmation, and packing and sorter input. [1]
The plan is to verify not only repetitive single tasks but whether the system can autonomously handle logistics work with multiple processes linked together and respond to exceptions that arise during operations. That framing is company-reported via Chosun Biz, not a completed KPI scorecard. [1]
Six processes at three centres
At major logistics centres in Jincheon, Yangsan, and Icheon, the company will pilot six processes: 3D sorter operations; GTP station picking and placement; safety monitoring and patrol management; small-item consolidation and packing; repair-parts picking and inspection; and small and mid-size home-appliance single-parcel picking. [1]
That mix spans classic pick-pack work, inspection, and facility patrol—useful breadth if the hardware and physical-AI stack can actually generalise across them.
Why it matters
A multi-year KEIT/MOTIE selection with named industrial and research partners is a concrete “state-backed adoption” datapoint for logistics humanoids in Korea—distinct from pure venture demos. The Jincheon apparel pilot gives the programme a prior on-site foothold rather than a greenfield fantasy. [1]
What we don’t know
- Budget size for the KEIT award (not stated in the piece).
- Which humanoid hardware platform will be used beyond the prior Jincheon bipedal pilot.
- Measured throughput, exception rates, or safety incident baselines versus human crews.
- How “domestic physical AI” will be evaluated against imported models in practice.
What this does not prove
Project selection does not prove that chained multi-process humanoid autonomy already works at Yangsan or Icheon, or that foreign-robot dependence will fall by a stated percentage. It proves Lotte and partners won a roughly five-year development-and-pilot window. [1]
The Bottom Line
Lotte Global Logistics has a roughly five-year national project window to prove domestic physical-AI humanoids can chain real logistics tasks—not just one pick cycle—across three Korean centres. Selection is real; production-scale autonomy is still the question the pilots must answer.
Sources
- Chosun Biz — 7 October 2026 — https://biz.chosun.com/en/en-industry/2026/10/07/FAXNSORGJBFMRP22BLY7GZMKYU/
