Strategy to gather data

Airborne Oil & Gas (AOG) is the world's first and leading manufacturer of composite pipes for the oil and gas industry. They had the ambition to largely automate the production environment. Landscape assessed the current IT ecosystem and helped developing a clear data strategy.

Having recently split-off from its parent company, AOG was working on setting up an independent IT environment. This included ERP software to support the primary business, and collecting all data necessary to optimise the production environment.

While it was obvious that a new system with new requirements was needed, and we were asked to investigate exactly where the new needs differed from the old. In addition, AOG had (and still has) the ambition to largely automate their production plant, which of course came with its own set of (data) requirements.

Assessment of IT requirements

Landscape assessed the application needs for AOG's primary business processes. This involved interviewing business stakeholders, critically assessing the stakeholders' IT requirements, mapping the requirements for these processes to IT applications and data sources, and advising on improvements. Such an overview of the organisation's IT requirements and their connection to business processes proved to be a very effective tool for communication and decision making.

With our analysis and recommendations, AOG has set up its own IT department and procured new software.

Understanding data assets

Next, Landscape worked on the issue of what data to collect in order to largely automate the production hall. In order for an AI to be able to do this process on its own, it needs to be trained on a large dataset of examples of how to deal with certain scenarios. Some of this data was already available, some of it was not.

Landscape assessed the primary - existing and future - data assets in the production hall, including temperature sensors and video cameras. For each, we estimated how much data was going to be collected, which was more than was ever possible to store. We could determine how much of this data should be stored, based on the requirements of training an AI to complete the required tasks.

Landscape's analysis and recommendations have formed the basis for several programs to improve data practices and to use data more effectively.