Operational performance of light electric freight vehicles in the last mile: two Nordic case studies

1. Introduction

In urban areas with increasing traffic and limited space leads to conflicting demands for its use. One challenge is finding a balance between creating livable spaces for people and accommodating deliveries and services crucial to the area’s functioning. Simply investing in more transportation infrastructure may not solve the issue. Thus, both industry leaders and academics are exploring alternative transportation solutions to mitigate the negative impacts of city logistics, with a focus on replacing traditional delivery vehicles with more sustainable options (Zhu et al., 2023). Many cities are experiencing densification and increased urbanization, which intensifies the competition for vacant areas and infrastructure among different road-user groups. Some cities in Europe are moving in the direction of more car-free centers with a focus on vibrant cities and facilities for soft road users (Hedblom et al., 2017). The negative impact of transport is not only focused on climate change but also has developed into a broader perspective including health issues (air quality and noise nuisance), public space occupancy and the attractiveness of cities in general (Hogt et al., 2017). Delivery of goods, on the other hand, has traditionally been adapted to a car-oriented infrastructure, primarily based on distribution solutions using diesel vans and trucks (Arnold et al., 2018).

The number of light commercial vehicles is growing in cities (Hogt et al., 2017). The demand for more goods, fast deliveries and an on-demand economy (Dablanc et al., 2017) in combination with the public sector zero-emission goals results in small deliveries and decreasing average shipment size, which in turn, contributes to an increase in the number of vans (Moolenburgh et al., 2020) and results in vehicles using less of their maximum capacity (Hogt et al., 2017). Increased e-commerce will continue to have a significant effect on in cities and contribute to a growing number of freight vehicles performing transport assignments (Visser et al., 2014; Moolenburgh et al., 2020). The increased demand for goods delivery in cities combined with a greater focus on access restrictions, environmentally friendly and vibrant cities accelerates the need to rethink traditional goods delivery operations. With increased demand for goods comes a corresponding need for space devoted to logistic activities.

The trends observed in cities put the logistics sector in a challenging situation where they are pressured to quickly lower emissions while also reducing their use of space in dense cities. One would expect the density of potential customers within cities to be a benefit for logistics companies, but operations can be hindered by congestion, significantly reducing efficiency. Logistics facilities closer to customer areas can allow the use of LEFVs to navigate dense areas more efficiently, but those facilities are expensive and demand dramatic increases in efficiency to offset costs. This phenomenon has been noticed by other researchers, recently by Buldeo Rai et al. (2022. p. 57) and Kin et al. (2024, p. 29), and we suggest referring to it as the “densification paradox” for logistics, where a concentration of customers does not necessarily lead to an increase in efficiency.

The aim of the paper is to provide operational insight into the use of cargo cycles using delivery data from two case studies plus interviews, showing under what circumstances they can be a useful complement to vans. The costs of the transshipment facilities were not included in the comparison as the necessary data were not available. Additionally, differences in the case studies in terms of scale, the use of subsidies, local context and the types of goods would make such a comparison of limited use. To use the framework of He (2020), we look at a single tier network for last mile delivery. Results indicate that cargo cycles are best suited to carrying smaller packages in areas that are difficult for larger vehicles to access with a relatively high customer density and a high number of stops. We also address some drawbacks and challenges in using cargo cycles effectively. In the following sections, we present a literature review, followed by the methods used and presentation of case studies. After that the results are presented and the paper closes with discussion, conclusions and future research.

2. Literature review

The last mile is well known to be the most expensive (Goodman, 2005), most complex (European Commission, 2011) and most polluting (Albergel et al., 2006) part of the logistic chain, with estimates suggesting it accounts from 28% (Ranieri et al., 2018) to more than 50% (Ostermeier et al., 2022) of the cost. Light electric freight vehicles (LEFVs) might be a step in the right direction to these challenges (Moolenburgh et al., 2020). This paper defines LEFVs in accordance with van Amstel et al. (2018) as a “bike, moped or compact vehicle with electric assistance or drive mechanism, designed for the distribution of goods in public space with limited speed. LEFVs are quiet, agile and emission-free and take up less space than conventional vans and trucks” (van Amstel et al., 2018, p. 4) In several European cities, urban planners are shifting their focus from automobiles to cyclists and pedestrians, which has contributed to cargo cycles becoming an alternative to motorized vehicles for urban freight (Dybdalen and Ryeng, 2021). Electric cargo cycles can play a key role as zero-emission vehicles in dense cities where the focus is on minimizing car use (Assmann et al., 2019). Cargo cycle distribution of packages and small shipments in central urban areas has great potential as cargo cycles are not impeded with the same restrictions related to queuing and parking as vans and trucks are. Cargo cycles often have the option of using bicycle lanes, roads and sidewalks, though the type of infrastructure used and where the cargo cycle parks can be influenced by the available infrastructure and its design. At the same time, the unfettered access of cargo cycles to sidewalks and other pedestrian areas can lead to safety concerns and other externalities, especially if cargo cycles are more widely adopted (Dalla Chiara et al., 2023). So far, these potential safety concerns have yet to materialize as no fatal accidents have been reported for cargo cycles and just a few injuries to riders and the public (van Duin, 2022).

In addition, cargo cycles avoid access restrictions in the form of zero-emission zones, tolls and other climate and environmental measures and can more easily maneuver away from unforeseen detours and closed roads (Sheth et al., 2019; Gruber et al., 2014). In Norway, and to some extent Sweden, bicycles can also ride on one-way streets where this is signed, which further increases the accessibility and flexibility of the cargo cycle compared to vans and trucks. However, cargo cycles are also limited by lower capacity (weight/volume), more limited range due to lower speeds and the added expense of transshipment operations and facilities. Therefore, even if delivery of goods with cargo cycles is considered an attractive solution in compact future cities, cargo cycles will not be suitable for all product segments and for already consolidated goods (Nordby Rundberget et al., 2019) without innovation. Many solutions require the implementation of two echelon vehicle routing problems that ensure the cycles are supplied with goods throughout the day either through mobile or fixed depots, but this adds complexity which adds to costs (Boysen et al., 2021).

The estimated potential for these small vehicles varies from being able to deliver from 10% (Melo and Baptista, 2017) to 25% (Wrighton & Reiter, 2016) of the total number of goods in cities. Wrighton and Reiter’s (2016) estimate rises to 50% if private citizens’ logistic trips are also included. Melo and Baptista (2017) used a traffic simulation model in AIMSUN 8.1.2 (Advanced Interactive Microscopic Simulator for Urban and Non-Urban Networks) to assess whether LEFVs, such as cargo cycles, could achieve better environmental and social results and still obtain the levels of operational and traffic efficiency for the city of Porto (Portugal). Robichet et al. (2022) studied DB Schenker’s operations in Paris and found that while it was theoretically feasible to distribute as much as 91% of their goods by cargo cycle using a weight limit of 200 kg, the costs would be prohibitive, with access to microhubs being the primary expense driver. Their estimate does not consider the volume of goods.

One of the biggest competitive advantages a cargo bicycle has over a delivery truck is its ability to quickly find parking close to the customer. A recent study conducted by INRIX in 2017 indicates that vehicles spend 9 min on average looking for on-street parking and this significantly adds to traffic congestion and emissions while Dalla Chiara and Goodchild (2020) found cruising for parking accounted for 28% of total trip time in Seattle on average. Trucks circling street blocks until parking is found hampers efficiency, adds to congestion and increases air pollution (Sheth et al., 2019).

Parsmo (2019) performed a lighter version of an LCA on propulsion for the use of cargo cycles in relation to electric and diesel vans. Not surprisingly, from energy efficiency perspective cargo cycles do well. The energy usage per kilometer for propulsion compared to an electric van decreases 7–12 times. In comparison to a diesel van, the decrease is 13–27 times. In both instances the truck was lightly loaded (50–200 kg) (Parsmo, 2019).

In Frankfurt and Utrecht, DHL has piloted a City Hub concept. A van drives into the city with multimodal ready containers that can be easily transshipped from the van to a cargo cycle. According to John Pearson, CEO DHL Express Europe, the cargo cycles offer a range of advantages; bypass congestion and make up to two times as many stops per hour than a van. The total cost of ownership is less than half than a van. But most importantly, there are no emissions (DHL Press release, 2017). In another study, Verlinghieri et al. (2021) similar benefits can be found with freight cycles, including lower costs, better maneuverability, less road space occupied and quicker delivery. By reducing delivery distances and times, cargo cycles can increase the efficiency of urban freight. Additionally, these businesses are aware that additional handling expenses in regional hubs can be reduced by employing intelligent procedures like containerization, standardization and automation (van Duin et al., 2022). Van Duin also has an updated scientific literature compilation on the topic of LEFVs. As an example, only one real life case has been found comparing vans to cargo cycles, Browne et al. (2011). They thoroughly investigated the case of Gnewt Cargo in London. From the company’s standpoint, the LEFV trial was a success in terms of transportation, the environment and finance because using LEFVs reduced total distance traveled and CO2 emissions per parcel delivered by 14 and 55%, respectively. After the project was completed, they made the decision to keep using LEFVs for delivery. This paper tries to improve this gap in previous literature, mentioned by van Duin (2022).

However, it is also clear that cargo cycles can only replace some activities within a transport system and effectively implementing them in cities is reliant on operational, vehicular, infrastructural, workforce, organizational and policy factors and a mixed fleet is likely to be the most efficient (Narayanan and Antoniou, 2022). These different factors can act as enablers or barriers for cargo cycles.

3. Methods and case description

This paper is centered on two case studies introducing cargo cycles in parcel delivery operations from Norway and Sweden, adding cargo cycles in one case and replacing two vans with two cargo cycles in the other case. The case of the city hub for Company A in Norway was the first of its kind to ensure zero emission deliveries within the entire city and offered perspective on the use of cargo cycles in cooperation with vans. The data from Company B provided a unique opportunity to compare cargo cycles and vans in operation. In both cases the data were provided by the companies.

Additionally, to understand the introduction of LEFVs in a wider context and to confirm common themes across the use of LEFVs, interviews with a Swedish cargo cycle producer and a Norwegian logistics company were conducted. The type of data sources used in this article comprises delivery data, interviews, email exchanges, company web sites and company documents as summarized in Table 1. The delivery data for each case study is described in greater depth in sections 4.5 and 4.6.

4. Results

This section presents the results from the in-depth interviews. The data of the people interviewed can be viewed in Table 1. The section ends with the presentation of the route data for the two cases as well as calculations.

5. Conclusion

For LEFVs to be seen as viable by LSPs they need to be considered the most efficient option. We see that companies are adapting to the new landscape within urban logistics by developing new KPIs to better understand the efficiency of these vehicles. Measures such as stops per route and joules per shipment introduce a time and energy component to measuring efficiency. This is in contrast to the long-haul logistics industry which focuses more on ton kilometers (tkm) and vehicle kilometers (vkm), as vehicles are likely to “cube out” or “weigh out” whereas in last mile distribution vehicles will often “time out”. Meaning that for parcels, oftentimes the driver’s workday is the limiting factor, not the weight or the volume of goods. The energy component stems from the vision/targets of the companies as well as a correlation between reduction in energy and cost savings.

The limitations of tkm and vkm for measuring the efficiency of lighter vehicles are illustrated in a report from a Swedish government agency (Trafa, 2023, Figure 1 and 2, page 13), which compared light goods vehicles (LGVs) with heavy goods vehicles in Sweden. They found that the LGVs moved 3% of the total freight tonnes and 2% of the total tkm in Sweden, but 70% of the mileage. LGVs also caused 50% of the NOx, about 50% of the particle matter, and 30% of the CO2. LGVs mostly operate in cities.

Similar to the findings from Browne et al. (2011), our case studies showed that cargo cycles drive fewer km than vans for similar routes. For Company B, the cargo cycles traveled between 23 and 32% fewer kms than vans in the delivery area, compared with 20% less in Browne et al.’s (2011) study. For Company A the difference was more extreme, with the cargo cycles traveling 44% fewer km on average.

After the initial journey from the terminal is controlled for, cargo cycles from Company B travel fewer kilometers than vans. Since the delivery area is the same and the cargo cycle delivers to more customers, the only plausible explanation is that the cycles have better access by avoiding traffic and using various forms of infrastructure such as roads, bike lanes and pedestrian walkways, as was frequently mentioned as a benefit in interviews. The cycles do not have to cruise for parking, can take short cuts and are not affected by one-way streets or traffic to the same extent as vans. Driving speeds in cities are generally low for both bikes and vans and many last mile operators spend more time per customer outside the vehicle delivering parcels by walking than driving (Dalla Chiara et al., 2023). Being able to use multiple forms of infrastructure leads to performance enhancements as the courier gets closer to the customer, faster and walks less.

One of the major differences between the two case studies looked at is that Company B directly replaced van routes with cargo cycles, whereas Company A incorporated them as part of a mixed fleet operating out of a larger city hub. The cooperation between the cargo cycles and vans makes it more challenging to isolate the performance of either the vans or the bikes, as they were reliant on each other and acted in cooperation to deliver as many goods as possible. Company A’s cargo cycles also carried substantially more weight and volume than Company B’s, though still not as much as the delivery vans. Proximity to the city hub and the ability to replenish goods throughout the day were reported to mitigate the limitations of cargo cycles related to capacity, allowing the cargo cycles to deliver greater volumes and weights.

Interestingly, the Browne et al. (2011) study also showed results related to distance traveled per parcel that we could not replicate with our available data. The total distance traveled per parcel was higher for cargo cycles in the Browne et al. (2011) study and the authors explain this by the limited load of the cargo cycles compared to vans as well as strict time requirements on some deliveries leading to less efficient routes.

The interviews confirmed the findings from literature; the running costs of cargo cycles are about half compared to vans and the purchasing costs are around a third. But in many instances the operational setups require some form of transshipment facility, which can be costly. Apart from the extra handling, hubs located centrally in cities are expensive and makes the creation of a profitable business model for cargo cycles challenging. While cargo cycles are less expensive to purchase and operate in terms of fuel/energy, maintenance issues remain a concern. In our case studies, and especially for Company A, the cargo cycles reportedly experienced significant challenges related to maintenance and parts availability.

Our work also supports previous research on bicycles traveling faster than cars in cities [1] by showing data that cargo cycles seem to mimic this pattern in relation to vans in city centers. It seems that cargo cycles deliver more efficiently than vans (measured by number of deliveries), at least with a customer density/route efficiency of over 2–3 customers per driven kilometer and for packages preferably of weights no larger than 5 kilos on average. Cargo cycles not only occupy less space, measured as fewer square meters per time unit but also drive fewer kilometers in total than vans.

A clear limitation of our study is that the additional costs related to the operation of hubs are not explicitly considered. Perhaps looking into a simpler form of a “terminal”, microhub or meeting points for transloading between vehicles is a way forward, to use infrastructure that is available at least some of the time in a day. Like parking lots in centrally located malls or gas stations that most likely are looking into new business models for the use of their space. The Department of Transportation in New York City are currently testing a version of this, dedicating curb side for delivery operations, see NYC DOT (2023).

Many cities have begun to enact restrictions on the movement of vehicles in an effort to combat issues related to congestion, local pollution and safety. Doing so also encourages the use of LEFVs, making them comparably more efficient than larger vehicles. Cities will need to consider how access to dense areas is maintained while creating pressure to use light vehicles such as cargo cycles, electric mopeds or looking further into the future, autonomous delivery vehicles.

Introducing cargo cycles into logistic systems creates both costs and benefits that are split between different actors. The logistics companies bear the extra cost of transshipment facilities and maintenance for instance, while cities and inhabitants benefit from less vehicle movement, less pollution and less congestion. As seen from the data from Company A, vans perform better than cargo cycles where the parcels are about double the size on average and the customers are further apart compared to Company B. The limited cargo space and larger parcels add several replenishment trips for Company A’s case.

Also, in Company B’s 2020 data, where the operational efficiency is measured in a 1:1 replacement of vans with cargo cycles, still some of the heavy parcels are delivered by a sweeper van.

Conventional diesel vans will slowly be replaced by electric vans, and these will be the preferred solution for most parcel delivery companies as they require fewer changes in how logistic systems are organized. However, swapping out fossil vehicles with electric ones does not address problems related to congestion, parking, safety or particle emissions from tires and brakes (not to mention the resources necessary for vehicle production). Research suggests (Melo and Baptista, 2017; Wrighton and Reiter, 2016) that between 10 and 50% of logistics activities (including private logistics) can be carried out by LEFVs, which could significantly reduce the negative impacts of logistics.

Cargo cycles offer a number of benefits but also carry some serious drawbacks. While cargo cycles and other LEFVs have the potential to deliver a higher proportion of goods, a number of barriers have so far hindered their transition from niche to mainstream. If LEFVs are adopted more widely, issues such as competition for space on sidewalks and bike lanes as well as regulation addressing the use of infrastructure for LEFVs could become more prevalent.

6. Future research

This work examines the use of LEFVs through two case studies. One in which the performance of cargo cycles is examined in relation to their cooperation with vans operating in the same area and a second case study, which offers a more direct comparison as cargo cycles are used to replace van routes in a specific area. Both studies show that cargo cycles can be competitive with vans if they are able to leverage the relative advantages they have through use of well placed, cost-efficient and organized transshipment points.

The paper adds valuable information for organizations that are thinking of trying LEFVs in operations as well as municipalities/local authorities that are interested in implementing more sustainable last mile deliveries. It is also important to mention that with the advent of AI in combination with automation other types of LEFVs are on the horizon, for instance slow moving side-walk delivery robots in suburbs where the competition for space is lower.

From the interviews, we know that the results can be very context dependent, and it would be of interest to understand how factors such as customer density/route efficiency, business models, package size, topography and congestion levels impact the efficiency of LEFVs.