In order to promote the sustainability and economic viability of Smart Local Energy Systems (SLES), energy products and services will need to change and adapt to new technologies, new tools and new techniques in order to be successful.

This report, Data Ethics and Bias: Practical steps to avoid discrimination in future Smart Local Energy Systems, demonstrates through a series of relevant use cases that data bias can create negative or discriminatory impacts on specific groups of users.

The increased uptake of controllable and smart technologies mean that responsible and ethical utilisation of personal data, and how it interacts with people, will quickly become a ubiquitous problem if safeguards and guidelines are not put in place.

A set of principles are proposed to prevent structural bias from entering the data that underpins these new products and services. These principles are partly drawn from existing work with use cases to support them, but also contextualised by Energy Systems Catapult to focus on energy.

Key points

This report is designed to support those implementing SLES to tackle the practical ethical challenges surrounding data when:

• Assessing risk in data collection
• Building algorithmic control systems
• Applying machine learning techniques.

To help with the practical ethical challenges of using data, we have set out three principles for addressing data bias in smart local energy systems.

• User group principles: a simplified method of identifying for whom the solution works and who is likely to have been excluded.
• User effect principles: a guide to negative impacts any bias may create for the users.
• Data mitigation principles: a guide to start de-risking data collection methodology to prevent bias from flowing through the process.

The principles are also contextualised in terms of smart local energy systems since they will be the focus of many of the more sensitive elements of granular and potential personal data. Guidance is given as to how to apply the data mitigation principles to promote good outcomes for consumers.