【正文】 nals perform well (above 35gmph under peak circumstances), and not in real life? This question we have asked ourselves, also to critically review our simulation models. In order to do so, we started from one of the current stateof the art fully automated facilities, and added latest improvements to the model to see whether we could increase the performance to levels that we do not experience in practice (yet). We used TBA‘s own proven container terminal simulation suite TimeSquare to quantify the effects of each adjustment individually. In this article we describe this stepwise improvement approach from an imaginary existing terminal with Dual RMGs and AGVs, as would have been constructed in the 1990s. For each step towards a stateoftheart terminal with TwinRMGs and LiftAGVs we show the effect on productivity of the various involved equipment types. Starting scenario: a Year 2020 automated terminal Our starting terminal is a fictitious terminal with 16 double trolley quay cranes (backreach interchange, with platform between the legs) on a 1,500m quay. The yard consists of 35 stack modules with dual crossover (or nested) RMGs. Crossover RMGs are stacking cranes that can pass each other (one is smaller and can pass the larger one underneath). Because of the passing ability, both RMGs are able to serve both the waterside and the landside transfer area in the perpendicular stack layout. Waterside transport is done by lifton liftoff (LOLO) Automated Guided Vehicles (AGVs), which are pooled over all quay cranes. All modeled equipment has technical specifications as is appropriate for 10yearold equipment. The terminal is suitable for a yearly throughput of million TEU (TEU factor )。 there is less than 5% transshipment. In peaks all 16 quay cranes will be deployed, and the peak gate volume equals 320 containers per hour. The yard can be stacked to fourhigh, and the peak yard density equals 85%. We have run an eighthour peak period with the simulation model to get the reference quay crane productivities of the starting scenario. The results are shown in Figure 1. In the remainder of the study we will specifically focus on a situation with five AGVs per QC (on average。 trucks picking up a container at the yard wait an additional two minutes (Figure 5). Figure 6 shows the status distribution of the RMGs, divided in RMGs processing the waterside (WS RMG) and RMGs processing the landside (LS RMG). Although they are dedicated to do productive moves of their corresponding side, they can do unproductive moves for either side. This is why the WS RMGs show a large increase in ?shuffle move‘ status when they execute shuffles for the gate moves. This takes the stress off landside RMGs that need to handle more trucks. In future steps we will see whether the waterside volume can be increased as well. Step 3: replacing AGVs by LiftAGVs LOLO AGVs require a ?handshake‘ interchange with RMGs at the yard. This causes waiting times for both RMGs and AGVs,because for almost every move one of them has to wait for the other to arrive. This handshake can be excluded from the process by using LiftAGVs instead of AGVs. LiftAGVs are able to place and take containers from a platform located in front of the stack modules by using a lift mechanism. RMGs place and take containers from the platform as well. In this step we use LiftAGVs with – besides the lifting ability –the same specs as the 10yearold AGVs. Changes and expected effects: Unlinked interchange between LiftAGV and RMG reduces waiting time for both equipments. This should increase overall terminal productivity. LiftAGVs need to make an additional stop in front of the container rack to lower or hoist their platform. This is an extra move in their routing process and costs additional time (15 – 25seconds per stack visit). This decreases productivity. The container racks require more space than interchange positions for AGVs. Therefore only four racks fit in each stack module interchange zone instead of five parking slots for reduces flexibility and has a negative effect on performance. Results The quay crane performance increases with 3 to bx/hr for any number of vehicles per crane. The reduced waiting times largely outweigh the longer drive times and fewer transfer points,as shown in Figure 7. Figure 8 shows the move duration per box of the AGVs and liftAGVs. In the left column for AGVs you can see a large portion of the time is consumed by ?Interchanging at RMG TP‘, minutes per box, which represents the waiting time for the handshake with an RMG. The right column for LiftAGVs shows a slight increase in dr iving times (because dr iving requires an additional action: lifting in front of rack), but also a huge reduction in ?Interchanging at RMG TP‘: only minutes (20 seconds).LiftAGVs are approaching quay cranes generally a bit earlier now, which causes ?Waiting for QC approach‘ to increase。 這些 加強 終端生產(chǎn)力 的解決方案是什么呢？ 在一 個與此 匹配 的小型 模擬 分析中 ， 對 所 有可能 影響船舶 生產(chǎn) 能 力 的因素進行 比較 。 為了 實現(xiàn)它 ，我們 以一個已有的完全 自動化 的系統(tǒng)設(shè)備為基礎(chǔ) ， 再增 加 最 新的 技術(shù) 改進 ，對于 沒有實踐經(jīng)驗 的我們，我們不知道 該模型 能 否 增加性能等級。 啟動 場景 ： 2020 年的 自動化終端 我們 剛 開始 是 在一個 1500 長的碼頭岸線上配有 16 臺雙梁軌道式碼頭橋式起重機 虛擬 的中轉(zhuǎn)站 （ 支腿移動的平臺是立體交叉道 ） 。 所有建模設(shè)備 都有相應(yīng)的 技術(shù)規(guī)格 ，其 使用年限為 10 年。 為了獲得可以參照的碼頭起重機生產(chǎn)能力啟動場景，我們已經(jīng)讓仿真模型運行了