档案库 · 物流与供应链 · 产品决策 · 2026
transload豪赌:CCTV能测量每一票零担货运:YC P26,HN上线,每站点每月约5万美元
三位慕尼黑工业大学的朋友将仓库安防摄像头变成货运体积测量仪,让零担承运商在不减慢装卸速度的情况下,捕捉低报费用的货物。
transload
做的是什么生意
A software company that turns the security cameras already installed in freight terminals into automatic dimensioners, measuring every LTL shipment as it moves through the normal dock workflow with no new hardware.
起因
The three founders - Nils, Julius and Jago - studied together at TU Munich and had known each other for eight years when they started transload in 2026. Jago grew up around the industry through his family's LTL trucking and cross-docking business. Their first idea was AI for optimizing forklift routes inside cross-dock terminals, but after talking to more than 50 trucking companies they kept hearing about freight dimensions instead - and realized that recent advances in monocular metric depth estimation made it possible to recover accurate 3D structure from ordinary camera footage without LiDAR.
经过
The company joined Y Combinator's Spring 2026 (P26) batch and launched publicly on Hacker News on June 8, 2026, then on Launch YC on June 21, 2026. Its two-step system connects a barcode scan timestamp to the correct shipment in video - reasoning over gaze, body orientation and movement - then segments the object and estimates a metric 3D bounding box, reporting under 1.5 inches of mean absolute error today. The founders say they are working with several LTL carriers, with one customer showing dimension errors on roughly 10% of checked shipments, and claim catching those differences is worth about $50K per site per month.
结果
Live with pilot customers and scaling the measurement itself: the team says its focus is pushing bounding-box accuracy below one inch at full coverage, with the first paid use case being revenue recovery - identifying under-dimensioned shipments and correcting billing - before moving on to trailer-utilization optimization.
背景
transload是一家旧金山初创公司,于2026年由三位慕尼黑工业大学的朋友创立——Nils(计算机视觉)、Julius(前麦肯锡)和Jago(工业传感器数据AI)——利用卡车货运终端中已有的安防摄像头测量零担货物。赌注是:承运商因按体积收费但信任货主申报的尺寸而损失数十亿美元,解决方案是软件,能在不增加装卸站的情况下默默测量每一票货物。
创始人的起点并非此想法。他们最初的概念是利用AI优化越库终端中的叉车路线,但在与50多家卡车运输公司交谈后,他们反复听到同一个痛点:货主低报货物尺寸。Jago通过家族的零担运输和越库业务在行业中长大。与此同时,空间AI刚跨过了一个门槛——单目公制深度估计大幅改进——使得从普通摄像头恢复3D结构成为可能,无需LiDAR。
transload的系统将每次条形码扫描与扫描时间戳附近视频中的正确物体关联,然后从单摄像头视图中拟合一个公制3D边界框;创始人报告目前平均绝对误差低于1.5英寸。终端已有固定摄像头、重复工作流程和条形码扫描,因此测量在后台进行,无需新硬件或改变装卸流程。其YC档案声称15%的货物实际尺寸大于申报,每站点每月在补费和水箱利用率方面价值约5万美元。
公司加入了Y Combinator 2026春季批次,于2026年6月8日在Hacker News上线(56分,19条评论),并于2026年6月21日在Launch YC上线,并表示正在与几家零担承运商合作,其中一家客户约10%的受检货物存在尺寸误差。其声明的首个用例是收入回收——识别尺寸不足的货物并附上视觉证据,以便承运商修正账单——然后再转向拖车利用率优化。
这件事要成立,得有什么
- 尺寸欺诈或错误是可测量的且重复发生:每票低报的货物都是损失的收入,因此客户可以在购买前计算每个终端的投资回报。
- 现有CCTV消除了采纳障碍:无需叉车行驶、装卸拥堵或新工作流程,这就是专用测量站只抽样检测的原因。
- 技术门槛刚刚移动:廉价单目公制深度使得从普通摄像头精确3D测量成为可能,无需LiDAR硬件。
- 狭窄的切入点保持专注:从低报的零担货物中回收收入是创始人在50多次卡车运输公司对话中验证的痛点。
可借鉴之处
选择客户在每次对话中都提到的痛点,然后利用他们已经拥有的资产:没有新硬件使得向终端销售变得可能。
后续进展
截至2026年9月2日,transload已有试点客户:其YC档案(2026春季批次)报告称15%的客户货物实际尺寸大于申报,每站点每月在补费和拖车利用率方面价值约5万美元,根据其Hacker News发布,该公司正与几家零担承运商合作,其中一家在受检货物中约10%存在尺寸误差。这支位于旧金山的三人物团队专注于测量精度,目标是在全覆盖下平均误差低于一英寸。
资料来源
- Launch HN: Transload (YC P26) - Measuring freight items with CCTV
- Launch YC: transload - we measure freight with security cameras
发现哪里写错了?告诉我们。
轮到你了
你刚读完一家。说说你在做什么,看看谁在赌同一件事。
免费账号 · 3 次免费提问 · 不用绑卡