The grocery delivery startup was considered "high-tech" because you could order groceries on a website and have them delivered. A darling of the late '90s dot-com boom, WebVan ultimately failed, becoming a poster child for the dot-com bust (the company went bankrupt in 2001) and a cautionary tale about the perils of so-called "last-mile" delivery -- the link in the supply chain that brings products to consumers.
Delivery is expensive and inefficient. Unlike most parts of the supply chain, the last mile doesn't scale, especially when a human employee drives a big car to bring one product to one consumer.
The "last mile" is ripe for disruption. But how?
Companies like Google and Amazon are trying. Google Express picks up goods at stores and delivers them to homes using route-optimization algorithms.
Google Express competes against Amazon's AmazonFresh grocery delivery service, which costs a whopping $299 per year -- and that's from a company famous for free delivery of regular packages.
(AmazonFresh is for groceries and should not be confused with recent reports that Amazon is working on its own shipping service for nongrocery packages to compete with FedEx and UPS.)
Peapod, Fresh Direct, Whole Foods and others are also trying to make grocery delivery work.
But none of these companies has really solved the last-mile delivery problem. It's still too slow and expensive. So what's the solution? Drones, right?
When Amazon CEO Jeff Bezos revealed his company's drone initiative on 60 minutes in 2013, the world was incredulous. Since then, Google and other tech companies, along with Wal-Mart and other retailers as well as numerous delivery companies like UPS, have unveiled drone-delivery experiments, plans and trials.
So the public can be forgiven for believing that the future of last-mile delivery is all about drones. But drones don't scale: They can't fly that far, can't carry that much and need to have their batteries recharged frequently.
While drones are great for bringing high-value items (like medicine) to remote places, they're not great for everyday deliveries.
The good news is that last-mile delivery can, in fact, scale. Several companies are already testing and showcasing technologies that could make that happen.
The rolling drone-base idea
Auto giant Mercedes-Benz recently rolled out a delivery concept that combines truck and drone delivery with robotic in-truck warehousing.
Mercedes' so-called Vision Van, which was created in partnership with a drone startup called Matternet, is an all-electric warehouse on wheels. Two internal racks, which are loaded with packages in the warehouse, are placed in the van. A robotic system grabs the right package based on the next delivery and hands it to the driver through a window while he's still in the driver's seat.
Alternatively, the robot system can push the package through a window in the ceiling and into the "arms" of one of two rooftop drones, which then delivers the package to the customer's backyard. This enables faster delivery to homes that are far from the road, and makes it possible for two or three packages to be delivered at once (one by the driver and two by the drones).
Another use for the drone port is to deliver a package to the driver that was ordered for same-day delivery after the driver left the warehouse.
The robotic in-van warehousing system increases efficiency and cuts down on errors, theoretically. The launching of the drones from vans, rather than from the warehouse, cuts way down on the battery problem.
Taken further, you could imagine the rolling drone-base idea leading to large trucks rolling into neighborhoods and then dispatching a fleet of drones, delivery bots and humans to deliver many packages at once.
The self-driving ice chest idea
An Estonian company called Starship Technologies makes a robot that looks like an ice chest on wheels for delivering packages and food autonomously or semi-autonomously.
Starship co-founder Ahti Heinla, who was one of the original creators of Skype, told me that his delivery robot is inherently safe because it's light and slow (4 mph) and has software that causes it to get out of the way and stop when pedestrians are close. It's also safe because the robot is supervised remotely by humans, who can take over if the robot can't manage. In 8,000 miles of testing, the robots have "encountered" more than 1 million pedestrians and haven’t injured anyone, according to Heinla.