Open Data (OD) is data available in a machine-readable format and without restrictions on the permissions for using or distributing it. OD may include textual artifacts, images, maps, video content, and other. The data can be published and maintained by different entities, both public and private. Despite its power to distribute knowledge freely and availability of a large number of datasets, OD initiatives face important challenges related to its widespread take up. More specifically, OD provisioning is based on a unidirectional linking from OD providers to OD users without considering requirements and preferences of the users. The OD users also lack metadata, and they need to develop specific technical solutions for providing a continuous OD flow and processing, which is particularly difficult when real-time OD are to be used. In this paper, we propose solving these challenges by envisioning a business ecosystem for OD. It is network-based, federated, and supports interplay between OD provisioning and knowledge management. As a methodological solution, we have applied the capability-driven development approach, which allows modeling of OD processing ecosystems, facilitates knowledge exchange about OD usage among members of the ecosystem, and supports configuring information systems for OD processing. The proposal is explicated with a theoretical study of its usability for the service of road maintenance in varying conditions.