Issue link: https://industrialengineer.epubxp.com/i/905872
December 2017 | ISE Magazine 31 dent Alvaro Gonzalez adapted and enhanced this model for the United States. Gonzalez used the city of Chicago (urban), the metropolitan area of Chicago (urban and suburban) and the state of Illinois (rural, suburban and urban) as the geographic areas for performing cost comparisons. The cost model esti- mates the average last-mile cost per delivered unit based on dis- tance, time, average quantity of products in the parcel, average number of stops per day per driver and extra handling costs. The costs are adjusted using coefficients for time windows, reverse logistics (including logistics handling and average handling time for putting returned items back in inventory), manned (signature delivery) or unmanned delivery, collec- tion points, pooling of parcels, type of vehicles, packaging and information and communications technology. These coeffi- cients were determined through discussions with three logis- tics providers in the Chicago area. Geographic and demographic data came from the U.S. Cen- sus Bureau. Trucking operation costs came from the American Transportation Research Institute. Data on extra travel time in big U.S. cities with densities between 1,200 and 1,800 inhabit- ants per square kilometer were obtained from studies done by INRIX and the Texas Transportation Institute of Texas A&M; University. Base reference case In order to compare last-mile costs across a range of options, the cost model has been run for a base reference case. This scenario assumes a single delivery route, no time windows, no secure boxes, no collection points, no returns, no advanced vehicle and routing technologies and no fuel-efficient vehicles. While the specific costs obtained are an approximation, the model provides valuable insights when evaluating differ- ent strategies relative to the base reference model. The base reference model for deliveries in the city of Chicago estimates average last-mile costs per unit delivered (or last-mile costs, for simplicity) as $2.66. This cost is indexed or represented as 100, and all other costs are scaled relative to this number. The aver- age cost for deliveries in the Chicago metropolitan area for the same base reference scenario is $3.10, a 16.54 percent premium to the city and an index of 117. For deliveries covering the entire state of Illinois, the average cost is $4.30, a 61.65 percent premium to the city and an index of 162, as shown in Figure 1. Effect of individual strategies Individual logistics strategies have varying effects on last-mile costs, often depending upon geographic region. For example, it is interesting to note that Figure 1 shows that a four-hour time window increases costs by more than 20 percent relative to a delivery with no time window. The cost of offering two-hour time windows in the city is cheaper than the base reference cost for deliveries across the entire state. This indicates that time windows could be offered in urban areas without a huge premium in costs. A city could promote the installation of secure boxes in every residential facility, including single-family homes and multiunit housing. This initiative might seem ineffective at first sight, but the results are surprising. For logistics companies it would represent great news as failed deliveries would disap- pear. As shown in Figure 1, the cost model suggests that this action alone would decrease the costs of delivery by around 25 percent when no time windows are offered. There may be more gains due to the possibility of more parcels per drop because customers might increase their order size. The public would benefit from a reduction in carbon dioxide emissions, lower city congestion and noise reduction. In a scenario where shippers deliver only to their customers' collection points, the company would be able to drop more parcels per stop. The primary limiting factor would be the space in the chosen vehicle and the associated load factors. The cost model simulation has been run for low and high load fac- tors. The results in Figure 1 show that last-mile costs can be reduced by nearly 50 percent with this strategy. Investing in routing tools can increase productivity through automating daily route planning, making real-time adjust- ments when needed and other features. The cost model simu- lations have been run for using routing tools for two options: FIGURE 1 Last-mile delivery strategies In general, adding time windows increases last-mile delivery costs, while using security boxes, collection points, routing tools, eco- friendly fleets and cargo bikes can save money. Collection points Routing tools Eco- friendly fleet AREA Base reference case 4-hour window 1-hour window No time windows 1-hour window Small load factor No time windows 20% fuel efficient No time windows 1-hour window Chicago 100 122 166 75 158 50 88 97 59 98 Chicago met 117 142 193 87 183 58 102 112 NA NA Illinois 162 197 268 121 255 81 140 154 NA NA The base reference cost for the city of Chicago is $2.66. This has been indexed as 100. Time windows Security boxes Cargo bikes