Reinforcement learning has previously been applied to the problem of controlling a perched landing manoeuvre for a custom sweep-wing aircraft. Previous work showed that the use of domain randomisation to train with atmospheric disturbances improved the real-world performance of the controllers, leading to increased reward. This paper builds on the previous project, investigating enhancements and modifications to the learning process to further improve performance, and reduce final state error. These changes include modifying the observation by adding information about the airspeed to the standard aircraft state vector, employing further domain randomisation of the simulator, optimising the underlying RL algorithm and network structure, and changing to a continuous action space. Simulated investigations identified hyperparameter optimisation as achieving the most significant increase in reward performance. Several test cases were explored to identify the best combination of enhancements. Flight testing was performed, comparing a baseline model against some of the best performing test cases from simulation. Generally, test cases that performed better than the baseline in simulation also performed better in the real world. However, flight tests also identified limitations with the current numerical model. For some models, the chosen policy performs well in simulation yet stalls prematurely in reality, a problem known as the reality gap.

Modern fast-growing broilers spend excessive periods resting and their activity further decreases with age. Inactivity has been suggested to increase impaired gait and the incidence of leg disorders. Tibial dyschondroplasia (TD) is a common leg pathology in broilers. A more complex environment might facilitate more activity and improve leg health. Perches or elevated platforms bring variety to broilers’ environment and could motivate more locomotion. This study examined the impact of perches and elevated platforms on walking ability, the occurrence of TD and level of bone ash and mineral contents. The investigation was performed on four commercial broiler farms throughout six consecutive batches with platforms and four to five with perches. On each farm at least two separate houses were included, enabling the comparison of furnished flocks to control flocks during each batch. Plastic slats with ramp access elevated by 30 cm or wooden perches of 10 and 30 cm height were offered in the furnished house. Farmers recorded the platform and perch usage twice a week with a five-point scale. Gait was scored before slaughter on a six-point scale according to the Welfare Quality® assessment protocol for poultry. The severity of TD was determined using a four-point scale on farm from all birds gait scored as 3 and at slaughter from 200 birds/flock. Farmers estimated 50% to 100% of the platforms to be occupied in all flocks throughout the entire growing period. Only single birds were perching, thus perch structures were constantly evaluated to be empty. Due to the low use, the perch-equipped houses were excluded when analysing bone content, walking ability and TD. On average, 30% of the tested birds exhibited gait score ⩾3. Younger scoring age resulted in a lower mean gait score and a lower percentage of scores 3 and 4 to 5. Overall, 2.3% of the birds examined at slaughter and 3.5% of the birds with gait score 3 were affected by TD. Leg health was better in birds with access to platforms: mean gait score, the percentage of birds scoring 3, and TD percentage and severity were lower in birds in platform-equipped houses. Elevated structures such as platforms, offering additional possibilities for locomotion to broilers seem to improve their leg health.

We review the behavioural needs, priorities and preferences of laying hens for increased space, perching, nesting, foraging and dustbathing behaviour. Hens make full use of and may work to gain access to perches, however it is not yet known how perching ranks in comparison with otherbehaviours. Laying hens appearto have an instinctive need to perform pre-laying (nest-building) behaviour and have a strong preference for a discrete, enclosed nest site, for which they will work hard to gain access as oviposition approaches. Access to a nest site is a high-ranking priority for laying hens, preferred over food at this time. Foraging is a behavioural need, with peat, sand and wood shavings preferred substrates in choice experiments. There is no reduction in time spent foraging when a cost is imposed, nor when feed is freely available. Dustbathing is currently viewed as a behavioural need, as the extent to which hens value dustbathing is not known. Bird preferences for space are complex and confounded by interactions between group size and stocking density. There is some evidence that priority for space varies during the day and increases when the total space available to a group of birds is restricted, and that greater priority is given to space than to small group size. The presence of apparently purposeless behaviour, of high levels of aggression or redirected behaviours such as feather pecking and cannibalism are indicators that the housing system is not meeting the behavioural needs of the hens and hence is not satisfactory forbird welfare.

Preliminary results are given from studies of cage-housed birds provided with various combinations of nest sites (of various designs), perches (placed at the front or back of the cages) and dust bath boxes. Tibial bone strength in end-of-lay hens was found to be markedly greater where perches were provided – suggesting a clear benefit from this addition. Complex nesting facilities (external nest boxes) elicited more normal pre-laying behaviour than nest sites within the cage. There appeared to be a need for two nests for group sizes of four or five birds.

The objective of the work described is the introduction of cage system modifications that enable the birds to express a wider range of behaviours, while preserving as far as possible the main advantages of the system. Preference testing and motivational studies have been undertaken in the process of assessing the value of increased space provision, perches and nest sites of various kinds.

Specific design features discussed are: the avoidance of traps which damage the plumage and cause injury, especially to feet, wings and head; the provision of abrasive strips on the cage baffle plate to keep claw length in check; the possible use of perches to improve foot condition and provide exercise; and the value of improved gate design, especially to reduce the risk of injury at end-of-lay bird removal. A new prototype multilevel system (Elson tiered terrace) is briefly described.