Using AI to find the best locations for urban logistics hubs
28 / 06 / 22 - 3 minute read
The ripple effects of COVID-19 and changing consumer habits are accelerating the shift towards e-commerce. This, in turn, is stimulating demand for last-mile logistics to cater for the growing ranks of online consumers demanding ove
Table of content
Where is best for a logistics hub?
So what is the ideal location for a logistics hub? Essentially, any facility should be near as many affluent consumers as possible. The rationale is that the willingness to pay for a logistics asset is predominantly a function of the expected urban freight turnover, that is, the distribution of goods to consumers.
Thanks to artificial intelligence (AI) and machine learning methods, the LIS computes the exact logistics catchment area that can be reached during a given travel time - for example, how far can be driven in 15, 30, 45 or 60 minutes. By combining
these geographical areas with socio-economic data, the LIS determines the number of households within the catchment area (the ‘demographic impact’) and their total spending power, that is, their economic impact. Finally, the LIS merges demographic and economic data to benchmark the best logistics locations within European cities.
The way and speed at which end-consumer goods are distributed within cities are re-shaping city logistics and the use of real estate assets.
Spending power counts in Berlin
PATRIZIA used the LIS to examine two European cities in depth: Berlin and London. The studies highlighted substantial differences between the capitals’ logistics networks. In Berlin, greater spending power and a denser distribution of households compensate for higher travel costs. In London, the purchasing power captured within shorter distances drives logistics rents the most. This illustrates the importance of understanding logistics in each urban context and how the LIS makes this possible.
With Berlin, the LIS examined the most active locations in terms of institutional transactions over the last two years. According to the Real Capital Analytics (RCA) database, there are three active areas inside Berlin: Berlin Tegel Airport, Treptow to the south of the city, and Neuenhagen to the east. The three most active areas outside the German capital are: Kloster Lehnin in the south-west, Großbeeren in the south and Grünheide in the south-east.
The six locations differ significantly in terms of economic and demographic catchment. On average, 169,000 households can be reached from the three inner-city areas within a 15-minute drive. By comparison, on average, the three outside locations capture less than 31,000 households. At this point, we could conclude that locations within the city centre are the strategic place to be to maximise the number of potential consumers.
Yet, interestingly, driving for longer outside the city appears to be well worth it. Driving 30 minutes from Großbeeren and Grünheide reaches far more homes than a 15-minute drive within the city, namely 4.7 times (795,000) and 2.1 times (354,000) more households respectively.
The data suggests that household economic power increases with distance from inner-city locations. Yet, outside the city, the opposite is true. For example, the average purchasing power per household within the 15-minute catchment area of Tegel Airport is €34,000 and rises by around 15% to €39,000 within the 60-minute catchment area.
Central locations best for London logistics
For London, PATRIZIA used Property Market Analysis (PMA) to identify well-established logistics submarkets to benchmark with the LIS, namely: Croydon, Enfield, Heathrow and Royal Park. Then it compared their historic rental growth for logistics.
Unlike in Berlin, the LIS reveals an inverse relationship between drive time and economic impact. For example, for Royal Park, close to the city centre, purchasing power per household falls as driving time increases.
The LIS proves that rises in demographic and economic impact need to compensate for greater travel times. For example, with Enfield, there is little increase in spending power per household between a driving time of 30 minutes (€65,000) and a 60-minute drive time (€67,000). However, there is a far more significant demographic impact.
Identifying unknown logistics locations
By linking LIS results to real estate performance indicators, PATRIZIA was able to gain significant insights into rental dynamics across the four London submarkets. PATRIZIA merged the CAGR of logistics rents and the economic impact of different travel times for each site. Results show that the most robust rental growth occurred in the locations ranking highest using the LIS tool, confirming its findings.
Across all of the economic impact areas, the15-minute drive time relates best to logistics rental growth. Moreover, the purchasing power captured within a 15-minute and 30-minute drive time has the biggest influence on growth in rental logistics property.
Essentially, the LIS algorithm can help investors comprehend the complexity of distribution networks across Europe and how patterns differ between cities. Moreover, the tool could be used to identify untapped urban logistics locations. And it could be
further tailored to meet the specific needs of logistics investment management teams.
Logistics is an adaptable sector with several subsectors, each of which can protect investors against demographic and economic changes. Site selection is a crucial component when acquiring outperforming assets. And the new AI tool has proven invaluable in predicting where suitable sites could be located, not only for current demand but also in the future.
We want to ensure the assets we invest in continue to perform, no matter what the future brings. Logistics and its multiple subsets, whether light urban industrial or major transportation hub or cold storage, are positioned to take advantage of long-term secular trends. It is a sector that was made for future-proofing a portfolio and as urban logistics grows in importance, so does the need to identify where precisely to locate a logistics asset.
Further Reading
