A variety of processes are executed in a typical warehouse, including receipt of goods, picking, packing, loading and transport. SiB Solutions wanted to develop with open-source technologies to avoid vendor lock-in. The company chose Apache Kafka, which is ideal for communication in the logistics environment, because it solves many issues inherent with large monolithic architectures, including a lack of reliability and flexibility. “[We] needed to have high volume streaming of data and Kafka is the best solution we could find for that,” Nilsson continues.
Event Streams is also integral to SiB Solutions’ Intelligent Video and AI Services, where Kafka is used to drive machine learning through analytics machines running neural networks. For SiB Solutions’ customers, video analytics are run onsite to comply with data residency regulations. Then edge computers process the videos onsite following a variety of deep learning models. The output of this process is sent from Kafka producers over Kafka topics to cloud servers, where application logic takes place. “[Kafka] allows us to build everything as microservices,” Nilsson adds.