#1111 2003 · Kiva Systems (Mick Mountz) · Warehouse robotics & logistics
Kiva sent robots to fetch the shelf instead of sending workers to walk to it
问题
Warehouse pickers walked miles a shift to fetch items, capping how fast and cheaply orders could ship
背景
Every conventional warehouse is built around the same assumption: inventory sits in fixed locations on shelves, and when an order needs an item, a human worker walks to that shelf, picks the item, and walks back — a layout inherited from decades before automation was practical. Mick Mountz saw the cost of that assumption up close working at Webvan, an online grocery delivery startup that collapsed in 2001 partly under the weight of expensive, labor-intensive warehouse fulfillment; a picker walking ten to fifteen miles a shift between the front of a warehouse and scattered inventory was slow, tiring and directly limited how many orders a facility could fulfill per hour.
The industry's standard fix was to optimize the walking: better warehouse layout, shorter aisles, smarter picking routes, more workers. None of those changes questioned the basic assumption that the person, being mobile and the shelf, being fixed, meant the person had to be the one who moved. Mountz, drawing on his MIT mechanical engineering training, asked why the relationship couldn't run the other way.
换别人会怎么做
Warehouse operators facing rising fulfillment costs and slow picking speeds typically responded by redesigning aisles for shorter walking distances, adding more pickers, or optimizing pick-path software — every fix assumed the human would keep walking to the shelf and tried to shrink the distance rather than eliminate it.
他们看到了什么
A shelf is just a payload, like an order box. Once a robot could carry either, there was no reason the expensive, slow-to-move resource — the worker — should be the one walking.
那一手
Mountz founded Kiva Systems in 2003 to build a fleet of squat, autonomous orange robots that slide under entire shelving pods, lift them, and carry the whole shelf to a stationary human picker standing at a fixed packing station — inverting decades of warehouse design so the inventory travels to the worker instead of the worker walking to the inventory, with software coordinating hundreds of robots to bring the right shelf to the right picker at the right moment.
为什么管用
A human picker's walking speed and fatigue set a hard ceiling on how many items a warehouse can fulfill per labor-hour, no matter how the aisles are arranged, because travel time scales with the distance between scattered inventory and the picker. Making the shelf mobile instead removes that ceiling: robots can travel far more densely and continuously than a human safely can, and concentrating all picking at fixed stations lets a single human handle a continuous stream of shelves without ever leaving their spot, converting a distributed walking problem into a centralized, robot-scheduled routing problem computers are much better at solving.
值了多少
Kiva sharply raised picking productivity; Amazon bought it for $775 million in March 2012, its second-largest deal ever.
什么时候会失灵
It requires a large enough, capital-intensive facility to justify the upfront cost of a robot fleet and the software to coordinate it, which rules it out for small warehouses or low-volume operations. It also depends on inventory being stored in units — pods, shelves, bins — that robots can physically lift and maneuver through aisles, which doesn't suit oversized, irregular or bulk-stored goods that don't fit the pod model.
后来呢
The goods-to-person warehouse model Kiva pioneered became the industry standard for large-scale e-commerce fulfillment, adopted or copied across the logistics sector well beyond Amazon, and reframed warehouse automation around moving inventory rather than moving people.
资料来源
- [1]Amazon Goes Robotic, Acquires Kiva Systems, Makers Of The Warehouse RobotSingularityHub, 2012singularityhub.com
- [2]Amazon buys robot maker Kiva Systems for $775 millionCNN Money, 2012money.cnn.com