Safety Case Workshop 2

Safety is a primary focus in the nuclear industry and is an integral feature of design, development, operations and decommissioning. In the UK, the Office for Nuclear Regulation (ONR) is the independent body which ensures that nuclear sites are suitably regulated and held accountable on behalf of the public. To this end, they require that nuclear site operators provide safety cases for their systems. A Safety Case is a structured argument that is used to provide evidence that a nuclear task system is safe. It considers aspects such as risks, failure modes, training needs, and reliability data. It is a concise and clear document, owned by the nuclear operator. Naturally, every nuclear site operator has a relationship with the ONR and, until recently, that relationship has been perceived as rather distant. Within RAIN we recognise the advantages of collaboration. We have arranged a series of workshops bringing together the ONR, nuclear operators, and academics. The first Safety Case Workshop was held in September last year. Details and presentations can be foundonline.

This second workshop, held in Manchester and facilitated by Matt Luckcuck and Michael Fisher from the University of Liverpool, focussed on the introduction of autonomous robotic systems into the nuclear industry. It was attended primarily by representatives from nuclear sites, as well as the nuclear supply chain, robotics industry, and academia.

Discussion was centred on four concrete case studies of robotics in the nuclear industry, which were used to explore the changes to safety verification and regulation with the advent of autonomy, machine learning, long-term usage, and so on.

The workshop aimed to:

• Build relationships between the supply chain, nuclear sites, the ONR, and academia
• Share concrete case studies of robotic systems from four different nuclear sites and discuss them in the context of safety
• Facilitate the open sharing of challenges and the collective discussion of solutions

The workshop was very well received.

An interested audience, coupled with enthusiastic presenters meant that it was very easy to understand the benefits of being at the session and to quickly determine the challenges (and their differences and similarities) at various nuclear sites.

Some of the key outputs from the workshop were;

• Relevant good practice for autonomy is still being developed
• The benefits of autonomy in the nuclear sector still needs to be justified
• It may be advantageous to invite non-nuclear representatives to future events to help diversify the points of reference for robots and autonomy
• A RAIN ‘principles’ document would be beneficial, to help frame thinking when considering robotic and autonomous systems
• Clear presentations of actual problems from industry are very useful for the supply chain

A follow up report, summarising the interesting discussions from the workshop, is being prepared.