Stefan Fedtke

An efficient part feeding is among the top challenges of many mass producers applying mixed-model assembly lines, for instance, in the automotive industry. This paper introduces a novel part feeding policy applied by a large German assembly plant for car engines: In-line kitting. Under this policy, the first stations of the line do not execute assembly operations, but are reserved for picking parts while passing containers of stock-keeping units (SKUs) arranged along the line. In this way, the parts are collected in traveling kits moving along with each workpiece on the conveyor, so that later assembly stations have the required parts directly available and do not lose precious labor time for unproductive parts handling. A major operational challenge when applying this part feeding policy is the walking effort for the human pickers while putting the SKUs of their respective stations into the traveling kits of the passing workpieces. Due to a high product variety, a large number of comparatively bulky SKU containers have to fit into each station, so that the walking distance to be covered by a worker during a work shift exceeds multiple kilometers. We show that this physical burden can be reduced significantly by balancing the workload among stations and optimizing the storage assignment of SKU containers within each in-line kitting station. We formulate the resulting optimization problem and provide suited solution procedures. Our computational study shows that the walking distance of pickers can be reduced significantly without producing any additional costs.

In the wake of e-commerce and its successful diffusion in most commercial activities, last-mile distribution causes more and more trouble in urban areas all around the globe. Growing parcel volumes to be delivered toward customer homes increase the number of delivery vans entering the city centers and thus add to congestion, pollution, and negative health impact. Therefore, it is anything but surprising that in recent years many novel delivery concepts on the last mile have been innovated. Among the most prominent are unmanned aerial vehicles (drones) and autonomous delivery robots taking over parcel delivery. This paper surveys established and novel last-mile concepts and puts special emphasis on the decision problems to be solved when setting up and operating each concept. To do so, we systematically record the alternative delivery concepts in a compact notation scheme, discuss the most important decision problems, and survey existing research on operations research methods solving these problems. Furthermore, we elaborate promising future research avenues.

Last mile deliveries with unmanned aerial vehicles (also denoted as drones) are seen as one promising idea to reduce excessive road traffic. To overcome the difficulties caused by the comparatively short operating ranges of drones, an innovative concept suggests to apply trucks as mobile landing and take‐off platforms. In this context, the paper on hand schedules the delivery to customers by drones for given truck routes. Given a fixed sequence of stops constituting a truck route and a set of customers to be supplied, we aim at a drone schedule (i.e., a set of trips each defining a drone's take‐off and landing stop and the customer serviced), such that all customers are supplied and the total duration of the delivery tour is minimized. We differentiate whether multiple drones or just a single one are placed on a truck and whether or not take‐off and landing stops have to be identical. We provide an analysis of computational complexity for each resulting subproblem, introduce efficient mixed‐integer programs, and compare all cases with regard to their potential of reducing the delivery effort on the last mile.

The distribution networks of the postal service industry are organized according to the hub-and-spoke paradigm, so that parcel distribution centers play a crucial role to consolidate the parcel flows to full truckloads. In these terminals, inbound trucks are unloaded at gates, shipments are identified, sorted by the central sortation conveyor system, and loaded into outbound trailers, in which they are moved toward their next destination. In this context, the scheduling of inbound trucks, which assigns a gate and a processing interval to each truck, is an essential operational decision problem. We formalize the resulting optimization problem and provide suited solution procedures. Furthermore, we test the impact of truck scheduling on the sortation performance of the central conveyor system with the help of a comprehensive terminal simulation. The online appendix is available at https://doi.org/10.1287/trsc.2016.0722 .

Fully automated sortation processes play a crucial role in modern distribution networks of the parcel service industry. In this context, this study systematically investigates different design alternatives of closed-loop tilt tray sortation conveyors. In particular, the number of loading stations for feeding parcels into a conveyor system, and its relation to the subordinate decision problem of assigning inbound and outbound destinations to docking doors, is shown to considerably influence the throughput of a hub terminal. A novel destination assignment problem is formalized, appropriate solution procedures are presented, and, by means of simulation, the resulting throughput is quantified, so that an experienced terminal manager can weigh the operational gains against the investment cost of different conveyor layouts.