A New Day for Home Delivery

ISE Magazine -Volume 48: Number 08
By Michael G. Kay

The rapid development of self-driving cars and aerial drones have made it seem likely that within a short time it will be feasible to have products delivered to the home without human intervention.

This industrial engineers think if the existing logistics network for retail products in the United States should be augmented with an automated delivery option from the existing set of stores and other retail establishments, or are there unique aspects of this new mode of delivery that warrant the reconsideration and possible redesign of the retail network to better realize the full potential of these automated delivery vehicles?

Logistics networks for home delivery companies like UPS and FedEx are designed to economize on their one big, indivisible cost of operation: the wage paid to the driver of each truck. To use the driver’s time best, a large-payload truck is used to make multiple deliveries along a carefully planned route. Of course, this makes it impossible to schedule deliveries for times that are convenient for a person at home, and often products are just left at the door.

More customized delivery

To support direct home delivery, a new type of logistics network is proposed that consists of a network of small distribution centers (DCs) located near the home. Goods would be delivered to the home in reusable standardized containers transported by driverless delivery vehicles.

In order to be located close to a home, the size of each DC has to be small, serving just several thousand people. In order to be cost-effective, the loading/unloading, sortation and storage capabilities of the DC should be automated fully, since each container delivered to a home might visit a dozen or more DCs while it’s in transit. In order for it to be economical to locate a fully automated DC close to a home, the capacity of the storage system should be able to be specified in small fixed-cost increments so that the cost of the DC is proportional to its size.

ISEs have the tools to design this future

The research needed to help realize the proposed home delivery logistics network is focused on the following additional basic technologies, all of which are within the scope of things done in industrial engineering:

  • Storage system control:The movement of containers across arrays of modules gives rise to a complex multiobject motion control problem that requires efficient algorithms.
  • Module design:Develop prototype modules and containers to determine the performance vs. cost trade-off with respect to container transit time across each module.
  • Network coordination:Develop a mechanism to coordinate the operation of each container, vehicle and DC in the network where, at each DC, containers going to the same DC compete to be in the next transported load, loads going to different DCs compete to be selected by a DDV, and DDVs are competing with each other to select loads for transport.
  • Performance analysis:Estimate delivery times and associated cost for a given logistics network.

The proposed delivery service is meant to be cost-effective even during the early stages of development for those like the elderly and disabled who would likely value the service even at a higher cost and longer delivery lead-time (initially, relatively few DCs would be available in the network).

Initially, the cost of this type of delivery service would only add to the cost of the goods purchased at the store (although it would eliminate the cost and time required for nonrecreational shopping).

Over time, this cost would decrease because stores would be able to eliminate many of their costs associated with stocking shelves and checkout, and they could implement a more efficient means of fulfilling orders, using more automated material handling equipment to receive goods and the same modular storage arrays used in the DCs to load orders directly onto driverless delivery vehicles for shipment to customers.

Existing retail stores would become, in effect, fulfillment warehouses, with possibly a front showroom area for customers to visit. And as the cost of a module decreases, many people might want to purchase a modular storage array for use inside their garage, especially since garage space would become available once people are able to call self-driving cars on demand for their transportation needs.