Search

German retail customers increasingly demand immediate, flexible, and sustainable deliveries in the urban area. Simultaneously, Germany is characterized by high urbanization, well-developed infrastructure, and a distinct online retail market. Consequently, crowdsourced delivery (CSD) by outsourcing the last-mile delivery to private couriers provides a promising and increasingly acknowledged delivery concept for German cities. Although existing literature has identified attributes that determine crowdshippers’ willingness to take on CSD, research studies lack the attributes’ relevance in combined consideration and their dependence on crowdshippers’ characteristics. A choice-based conjoint analysis with 193 German potential CSD crowdshippers was performed to examine the present research gap. The results reveal delivery time and remuneration as determinant attributes, whereas the degree of familiarity with the recipient and the parcel weight proved less important. In this way, this study is the first to demonstrate the diverging relevance of CSD attributes in the assignment of deliveries. Furthermore, this study proves that the attributes’ relevance is influenced by the crowdshippers’ surface-level and deep-level characteristics. By conducting a sensitivity analysis, this study is the first to determine how the increase in additional travel time affects the willingness to conduct a CSD. Moreover, the influence of the crowdshippers’ familiarity with the recipient on the overall results is low, indicating an application of the concept also used outside of family networks. For CSD platform operators, the results outline how to individualize CSD orders for active crowdshippers to increase their engagement. Policy makers will be informed about how CSD is an opportunity to make transport in the city more sustainable.

Maritime container shipping (MCS) firms represent the backbone of worldwide supply chains. Due to the emergence of global trends and disruptions, MCS firms increasingly face an uncertain environment. Consequently, MCS firms must develop dynamic capabilities to enable the reconfiguration of organizational resources for building resilience in a constantly changing business ecosystem. To create adequate dynamic capabilities, MCS firms need to anticipate the future of their macro environment through advanced foresight techniques. Using a Delphi-based scenario analysis, this study systematically examines scenarios for MCS firms’ macro environment. Twelve projections for the MCS industry were systematically created and were then assessed by 51 maritime experts. The resulting three distinct scenarios deliver valuable insights for MCS firms’ executives, whereas the blockchains and drones technology will have already increased efficiency in the short-term scenario “picking the low-hanging fruit,” other technological disruptions will only affect the industry in the medium-term scenario “experiencing an era of dichotomy.” In the long-term scenario “overcoming the obstacles of the generational shift,” MCS firms will experience a transfer toward alternative fuel powered and autonomously driven vessels. Offering detailed scenarios for the future MCS macro environment, this study represents a guide for decision-makers on how to create dynamic capabilities in MCS firms to build resilience. Moreover, the results revealed differences in the experts’ assessments due to their characteristics. Therefore, the study contributes to the academic focus on in-depth diversity analysis in the Delphi methodology and emphasizes the relevance of incorporating multiple stakeholders and panelists in future planning.

Neuromorphic computation-in-memory fabric based on emerging nonvolatile memories considered an attractive option to accelerate neural networks (NNs) in hardware as they provide high-performance, low-power, and reduced data movement. Although nonvolatile resistive memories (NVMs) offer many benefits, they are susceptible to data retention faults, where previously stored data are not retained after a certain amount of time due to external influence. These faults are more likely to happen unidirectional and severely impact the inference accuracy of the hardware implementation of NNs since the synaptic weights stored in the NVMs are subject to retention faults. In this work, we propose an approximate scrubbing technique for NVM-based neuromorphic fabric to mitigate unidirectional retention faults with virtually zero storage overhead depending on the definition of scrub area for multilayer perceptron (MLP) and convolutional NNs (CNNs). The training of the NNs is adjusted accordingly to meet the requirements of the proposed approximate scrubbing scheme. On different benchmarks, the proposed scrubbing approach can improve the inference accuracy up to 85.51% for MLP and 87.76% for CNN over the expected device operational time with negligible storage overhead.