AI-readiness and production resilience: empirical evidence from German manufacturing in times of the Covid-19 pandemic

TLDR

The Covid‑19 pandemic caused widespread production restrictions, underscoring the critical role of production resilience and suggesting that AI technologies can enhance firms’ ability to withstand and recover from such disruptions. The study examines whether AI‑readiness influenced German manufacturers’ production resilience during the spring 2020 lockdown. Using data from two large‑scale surveys of 237 German manufacturers, the authors performed quantitative analyses to test hypotheses about AI readiness and resilience. Results indicate that AI‑enabled production improves resilience during lockdowns, but only when firms surpass a certain AI threshold, offering both research insights and practical recommendations for resilience management.

Abstract

The outbreak of the Covid-19 pandemic led to restrictions in production worldwide. Numerous firms were affected and unable to keep up production due to lockdowns. In disruptive events like this, the resilience of the production system is of central importance, as the survivability of the entire firm depends on it. In this context, the literature argues that cutting-edge technologies, such as Artificial Intelligence (AI), raise the proactive and reactive capabilities of firms, enabling them to better resist and recover from disruptive events and thus, show a higher resilience. This paper takes up this topic and observes the Covid-19 pandemic with the aim to analyse whether a firm's AI-readiness had an impact on its production resilience during the spring 2020 lockdown in Germany. For this purpose, we combine two large-scale surveys containing data from 237 manufacturers in Germany and test hypotheses based on quantitative analyses. Our results show that firms could indeed benefit from AI-enabled production during the lockdown. However, it is also clear that manufacturers have to exceed a certain AI threshold to significantly increase their resilient capabilities and realise positive effects. Our findings not only hold implications for research, but also provide recommendations for the resilience management of manufacturers.