The investigation into the 2010 freight train roll-back incident at Shap looked at tools that assess rosters for fatigue. It found that different tools made different predictions about fatigue based on assumptions about whether people adjust to night shifts and experience less fatigue over time, or whether they are likely to become increasingly tired over consecutive days. If a person’s body clock is adjusted to being awake during the day and asleep at night they will struggle to stay alert during their first few night shifts. In addition, if a person adopts their normal daytime routine leading up to their first night, by the end of that shift they will have been awake for a very long time. And we know that when people have been awake for 17 hours they start to show signs of impairment that are similar to being over the drink-drive limit.