Processing of complex traffic scenes for effective steering and collision avoidance: a perspective, from research into human control, on the challenges for sensor-based autonomous vehicles on urban roads.

An overview is provided of behavioral research into human steering and collision avoidance including the processing of optic flow, optical looming and the role of the human mobile gaze system. A consideration is then made of the issues that may occur for autonomous vehicles (AV) when they move from grid-type road networks into complex inner-city streets and interact with human drivers, pedestrians and cyclists. Comparisons between human processing and AV processing of these interactions are made. This raises issues as to whether AV control systems need to mimic human visual processing more closely and highlights the need for AV systems to develop a "theory of road users" that allows attribution of intent to other drivers, cyclists or pedestrians. Guidelines for the development of a "theory of road users" for AVs are suggested.

The use of visually guided behaviour in children with Developmental Coordination Disorder (DCD) when crossing a virtual road.

The ability to safely cross a road is a perceptual-motor skill that involves coordination between a pedestrian's perception of the approaching vehicles and their locomotive capability to execute the road crossing action. Developmental Coordination Disorder (DCD) is a chronic disorder that is characterised by significant motor difficulties that impact on daily living, including a reduced ability to perform visually guided actions. A total of 25 typically developing primary school aged children and 25 age- and gender-matched children with DCD were presented with a virtual desktop task that required them to select suitable temporal crossing gaps between vehicles a stream of traffic approaching at either 20mph, 30mph or 40mph from the near-side (one-lane) or both near+far-sides (two-lane). A best-PEST staircase procedure was used to measure the temporal gaps that children accepted and the maximum likelihood value was taken after nine reversals as each participant's threshold. Typically developing children accepted temporal gaps that were sufficient to execute a safe crossing for vehicles approaching at 20mph and 30mph, but insufficient for vehicles approaching at 40mph. In contrast, children with DCD selected insufficient temporal crossing gaps across all approach speeds, which if translated to the roadside would have resulted in collision. These findings add to our understanding of the difficulties that children with DCD appear to have with visually guided behaviour and suggest the potential impact on one aspect of daily functioning that could have significant consequences.

Eye and hand movement strategies in older adults during a complex reaching task.

The kinematics of upper limb movements and the coordination of eye and hand movements are affected by ageing. These age differences are exacerbated when task difficulty is increased, but the exact nature of these differences remains to be established. We examined the performance of 12 older adults (mean age = 74) and 11 younger adults (mean age = 20) on a multi-phase prehension task. Participants had to reach for a target ball with their preferred hand, pick it up and place it in a tray, then reach for a second target ball and place that in the same tray. On half the trials (stabilising condition), participants were required to hold the tray just above the surface of the table with their non-preferred hand and keep it as still as possible. Hand and eye movements were recorded. Older adults took longer to complete their movements and reached lower peak velocities than the younger adults. Group differences were most apparent in the stabilising condition, suggesting that the added complexity had a greater effect on the performance of the older adults than the young. During pickup, older adults preferred to make an eye movement to the next target as soon as possible, but spent longer fixating the current target during placement, when accuracy requirements were higher. These latter observations suggest that older adults employed a task-dependent eye movement strategy, looking quickly to the next target to allow more time for planning and execution when possible, but fixating on their hand and successful placement of the ball when necessary.

Cortical functioning in children with developmental coordination disorder: a motor overflow study.

This study examined brain activation in children with developmental coordination disorder (DCD) to reveal areas that may contribute to poor movement execution and/or abundant motor overflow. Using functional magnetic resonance imaging, 13 boys with DCD (mean age = 9.6 years ±0.8) and 13 typically developing controls (mean age = 9.3 years ±0.6) were scanned performing two tasks (finger sequencing and hand clenching) with their dominant hand, while a four-finger motion sensor recorded contralateral motor overflow on their non-dominant hand. Despite displaying increased motor overflow on both functional tasks during scanning, there were no obvious activation deficits in the DCD group to explain the abundant motor overflow seen. However, children with DCD were found to display decreased activation in the left superior frontal gyrus on the finger-sequencing task, an area which plays an integral role in executive and spatially oriented processing. Decreased activation was also seen in the left inferior frontal gyrus, an area typically active during the observation and imitation of hand movements. Finally, increased activation in the right postcentral gyrus was seen in children with DCD, which may reflect increased reliance on somatosensory information during the execution of complex fine motor tasks.

Optimal use of visual information in adolescents and young adults with developmental coordination disorder.

Recent reports offer contrasting views on whether or not the use of online visual control is impaired in individuals with developmental coordination disorder (DCD). This study explored the optimal temporal basis for processing and using visual information in adolescents and young adults with DCD. Participants were 22 adolescents and young adults (12 males and 10 females; M = 19 years, SD = 3). Half had been diagnosed with DCD as children and still performed poorly on the movement assessment battery for children (DCD group; n = 11), and half reported typical development (TD group; n = 11) and were age- and gender-matched with the DCD group. We used performance on a steering task as a measure of information processing and examined the use of advance visual information. The conditions varied the duration of advance visual information: 125, 250, 500, 750, and 1,000 ms. With increased duration of advance visual information, the TD group showed a pattern of linear improvement. For the DCD group, however, the pattern was best described by a U-curve where optimal performance occurred with about 750 ms of advance information. The results suggest that the DCD group has an underlying preference for immediate online processing of visual information. The exact timing for optimal online control may depend crucially on the task, but too much advance information is detrimental to performance.

A representation of changing heading direction in human cortical areas pVIP and CSv.

When we move around in the environment, we continually change direction. Much work has examined how the brain extracts instantaneous direction of heading from optic flow but how changes in heading are encoded is unknown. Change could simply be inferred cognitively from successive instantaneous heading values, but we hypothesize that heading change is represented as a low-level signal that feeds into motor control with minimal need for attention or cognition. To test this, we first used functional MRI to measure activity in several predefined visual areas previously associated with processing optic flow (hMST, hV6, pVIP, and CSv) while participants viewed flow that simulated either constant heading or changing heading. We then trained a support vector machine (SVM) to distinguish the multivoxel activity pattern elicited by rightward versus leftward changes in heading direction. Some motion-sensitive visual cortical areas, including hMST, responded well to flow but did not appear to encode heading change. However, visual areas pVIP and, particularly, CSv responded with strong selectivity to changing flow and also allowed direction of heading change to be decoded. This suggests that these areas may construct a representation of heading change from instantaneous heading directions, permitting rapid and accurate preattentive detection and response to change.

Detection of vehicle approach in the presence of additional motion and simulated observer motion at road junctions.

One of the key contributory factors for accident involvement is the misjudgment of vehicle approach. Past research has indicated that individuals can use the rate of visual "looming" in order to judge the time to arrival (TTA) of approaching vehicles. Although a large number of road traffic collisions occur at roadside junctions, very little research has focused on individuals' abilities to detect the onset of visual looming within a complex road scene at junction scenarios. In this research, computer generated scenes with photorealistic vehicle images, and a psychophysical staircase methodology, were used to explore drivers' ability to detect the approach of both motorcycles and cars within a contextually rich city scene. Across three experiments the effect of additional vehicular and observer motion on driver detection of vehicle approach was assessed. Results showed that individuals were significantly poorer at detecting the approach of the motorcycle stimulus compared with the car stimulus. Results also showed that additional vehicular motion within the scene had a negative effect on detection thresholds for the car stimulus. Finally, the results showed that introducing lateral global motion of the scene, such as might occur if the observer was moving steadily forward from a junction, negatively affected detection thresholds. The theoretical implications of the findings are discussed, including how vehicles traveling at high speed are often below the threshold for detecting visual looming. Practical implications for road design and layout are discussed that address the perceptual errors noted.

Errors in motion processing amongst older drivers may increase accident risk.

Accident statistics highlight that older drivers are more frequently involved in right-of-way collisions than younger drivers. Accurately gauging vehicle speed is critical for judgement of when to pull out from a junction safely in front of oncoming traffic. We used psychophysical methods to measure drivers' ability to discriminate between different rates of looming presented by vehicles approaching at different speeds. We demonstrate that sensitivity to approach speed reduces by between 2.8 and 3.4 mph, dependent upon vehicle type, for every decade that age increases. We show that perceptual limitations for drivers over the age of 75 years can lead to a 50% reduction in time available to perform traffic manoeuvres, which may contribute in part to their overrepresentation in casualty statistics at junction. Results are discussed in terms of implications for road safety policy.

Obstacle avoidance and smooth trajectory control: neural areas highlighted during improved locomotor performance.

Visual control of locomotion typically involves both detection of current egomotion as well as anticipation of impending changes in trajectory. To determine if there are distinct neural systems involved in these aspects of steering control we used a slalom paradigm, which required participants to steer around objects in a computer simulated environment using a joystick. In some trials the whole slalom layout was visible (steering "preview" trials) so planning of the trajectory around future waypoints was possible, whereas in other trials the slalom course was only revealed one object at a time (steering "near" trials) so that future planning was restricted. In order to control for any differences in the motor requirements and visual properties between "preview" and "near" trials, we also interleaved control trials which replayed a participants' previous steering trials, with the task being to mimic the observed steering. Behavioral and fMRI results confirmed previous findings of superior parietal lobe (SPL) recruitment during steering trials, with a more extensive parietal and sensorimotor network during steering "preview" compared to steering "near" trials. Correlational analysis of fMRI data with respect to individual behavioral performance revealed that there was increased activation in the SPL in participants who exhibited smoother steering performance. These findings indicate that there is a role for the SPL in encoding path defining targets or obstacles during forward locomotion, which also provides a potential neural underpinning to explain improved steering performance on an individual basis.

Reaching a better understanding of the control of bimanual movements in older adults.

The ability to interact skilfully with the environment is essential for independent living and therefore a critical factor for the aging population. Here we investigate the differences between young and older adults in a bimanual reaching task where the goal is to bring two objects together to the same location with a synchronous placement. Older (mean age 74) and young (mean age 20) adults were asked to pick up two spatially disparate objects, one in each hand, and bring them together to place them in one of three trays laid out in front of them from left to right. The results showed that the older adults were no more detrimentally affected than the young by asymmetric bimanual movements compared to symmetric ones, and both groups completed their movements in the same time. Nevertheless, compared to the young, the older adult group produced reaches characterised by higher peak velocities (although this effect was marginal), shorter hover times, and where the movement distance varied for each hand the scaling of the kinematic profile across the two limbs diverged from that found with younger participants. They then spent longer than the young in the final adjustment phase and during this phase they made more adjustments than the young, and as a result were more synchronous in terms of the final placement of the objects. It seems that the older adults produced reach movements that were designed to reach the vicinity of the tray quite rapidly, after which time they made discreet adjustments to their initial trajectories in order to exercise the precision necessary to place the objects in the tray. These findings are consistent with the idea that older adults have problems using online control (as they wait until they can fixate both objects before making adjustments).

Reduced looming sensitivity in primary school children with Developmental Co-ordination Disorder.

Almost all locomotor animals are sensitive to optical expansion (visual looming) and for most animals this sensitivity is evident very early in their development. In humans there is evidence that responses to looming stimuli begin in the first 6 weeks of life, but here we demonstrate that as children become independent their perceptual acuity needs to be 50 to 100 times better than has been demonstrated in infants in order to be skilful at collision avoidance at a roadside. We have recently established that sensitivity to the detection of visual looming in 6- to 11-year-old children is significantly below that of adults (Wann, Poulter & Purcell, 2011). Here, using comparable methods, we explore looming detection sensitivity in children with Developmental Co-ordination Disorder (DCD), who show broad patterns of impairment in visuo-motor control. We presented visual simulations of approaching vehicles, scaled to represent different approach rates, to children with DCD aged between 6 and 11 years (n = 11) and typically developing age and gender matched controls (n = 11). Looming detection thresholds were measured under foveal and perifoveal viewing conditions, for isotropic expansion and isotropic expansion with simulated viewpoint motion. Our results show that there are situations in which children with DCD may fail to detect vehicles approaching at speeds in excess of 22 km/h, suggesting a developmental immaturity in looming sensitivity. This provides one of the first clear demonstrations of low-level motion processing deficits in children with DCD. The decrement observed may give rise to potential errors in the road crossing behaviour of these children, whereby approaching vehicles could be perceived as stationary. These findings further contribute towards understanding the adverse statistic that children under 9 years of age are four times more likely than adults to be involved in a road accident as a pedestrian.

Driving skills of young adults with developmental coordination disorder: Maintaining control and avoiding hazards.

In this study we assess for the first time the driving skills of young adults with developmental coordination disorder (DCD). We use a virtual city and a driving simulator to examine steering control, speed regulation and the responses to pedestrians on the road. Participants were adolescents and young adults who had been diagnosed with the disorder as children. For most participants the symptoms were maintained (DCD group) but for others they had largely dissipated (AD group). We also invited typically developing (control) participants matched in age, gender, and driving experience to the DCD and AD participants. Compared to their matched controls, the DCD group showed difficulties in steering when turning bends but not when driving along straight roads. Although the average speed of the DCD group was similar to their controls this may have been too fast for them to steer effectively around the bends. The DCD group also took 50% more time to react to pedestrians who walked towards their path. We found no such differences between the AD and their matched controls. We discuss the results in terms of visual information processing and suggest further applied and fundamental research on this topic.

The reliance on visual feedback control by older adults is highlighted in tasks requiring precise endpoint placement and precision grip.

There is an ongoing debate as to whether a greater degree of sensory-motor control is required to maintain skills as humans progress to be septuagenarians. Here, we investigate the dependence of older participants upon vision to execute skilled prehension movements. In a first experiment, participants were required to place a small, round peg in one of three randomly cued holes. A mirror apparatus was used to create conditions where they could always see the target locations, but vision of their hand approaching the target could be removed, and we explored end position accuracy. In a second experiment, we examined the ability of participants to precisely control their grasp action under conditions where they could see the objects but not their hands completing the action. The results showed that in Experiment 1, the older adults undershot the target in their primary submovement and hence had to move further in their secondary movement to achieve their goal. In Experiment 2, the older adults spent longer in the final adjustment phase (a near zero velocity phase at the end of the reach) when vision of the hand was unavailable. These findings suggest that older adults are indeed more reliant on visual feedback than the young in tasks that require precise manual control, and this clarifies conflicting accounts in the prior literature.

Reduced sensitivity to visual looming inflates the risk posed by speeding vehicles when children try to cross the road.

Almost all locomotor animals respond to visual looming or to discrete changes in optical size. The need to detect and process looming remains critically important for humans in everyday life. Road traffic statistics confirm that children up to 15 years old are overrepresented in pedestrian casualties. We demonstrate that, for a given pedestrian crossing time, vehicles traveling faster loom less than slower vehicles, which creates a dangerous illusion in which faster vehicles may be perceived as not approaching. Our results from perceptual tests of looming thresholds show strong developmental trends in sensitivity, such that children may not be able to detect vehicles approaching at speeds in excess of 20 mph. This creates a risk of injudicious road crossing in urban settings when traffic speeds are higher than 20 mph. The risk is exacerbated because vehicles moving faster than this speed are more likely to result in pedestrian fatalities.

Roadside judgments in children with Developmental Co-ordination Disorder.

As pedestrians, the perceptual ability to accurately judge the relative rate of approaching vehicles and select a suitable crossing gap requires sensitivity to looming. It also requires that crossing judgments are synchronized with motoric capabilities. Previous research has suggested that children with Developmental Co-ordination Disorder (DCD) may have deficits in visual processing, specifically in detecting visual motion. It is possible, therefore that this population are at greater risk at the roadside. In a series of motion prediction tasks, several component roadside skills were assessed in 15 children with DCD, or at risk of DCD, aged between 6 and 11 years along with 15 typically developing age and gender matched controls. First, threshold errors for relative approach rate judgments (looming) were measured when vehicle size (car or truck) varied. Second, thresholds for crossing gap selection were measured for vehicle approach speeds of 32, 48, 64 and 80 km/h (20-50 mph). These were related to the walking speeds of children of different ages and profiles. We found that children with DCD showed a deficit in making relative approach rate judgments, using looming, which suggests they may not discern that a vehicle is travelling faster than the urban speed limit. Children with DCD also left considerably longer temporal crossing gaps than controls perhaps reflecting a lack of confidence in their ability, these preferred gaps were over twice the average inter-car gaps that occurred on roads around their school. Our findings raise a number of issues concerning children with DCD and their competence and potential limitations as pedestrians.

Driving skills of young adults with developmental coordination disorder: regulating speed and coping with distraction.

In two experiments, we used an automatic car simulator to examine the steering control, speed regulation and response to hazards of young adults with developmental coordination disorder (DCD) and limited driving experience. In Experiment 1 participants either used the accelerator pedal to regulate their speed, or used the brake pedal when they needed to slow down from a pre-set speed. In Experiment 2, we introduced an auditory distraction condition that shared similarities with maintaining a conversation. Overall, the DCD group produced a larger variance in heading and needed more steering adjustments on straight roads, compared to age-matched controls. When turning bends, the DCD group showed greater difficulty in controlling steering while regulating their speed with the accelerator pedal but this was less problematic when using the brake. The DCD group also responded slower than the control group to pedestrians who walked towards their path. The auditory distraction in Experiment 2 had no visible effects on steering control but increased the reaction times to pedestrians in both groups. We discuss the results in terms of the visuomotor control in steering and the learning of optimal mappings between optic flow and vehicle control.

Neural processing of imminent collision in humans.

Detecting a looming object and its imminent collision is imperative to survival. For most humans, it is a fundamental aspect of daily activities such as driving, road crossing and participating in sport, yet little is known about how the brain both detects and responds to such stimuli. Here we use functional magnetic resonance imaging to assess neural response to looming stimuli in comparison with receding stimuli and motion-controlled static stimuli. We demonstrate for the first time that, in the human, the superior colliculus and the pulvinar nucleus of the thalamus respond to looming in addition to cortical regions associated with motor preparation. We also implicate the anterior insula in making timing computations for collision events.

Integration of dynamic information for visuomotor control in young adults with developmental coordination disorder.

We examined the hypothesis that developmental coordination disorder (DCD) consists of a poor integration of distal preparatory visual information with the visual information that arises during movement execution. We set up a steering task where the action goal was to steer smoothly on a virtual winding course under conditions that manipulated the availability and timing of visual information. Participants were 20 young adults who had been diagnosed with DCD in their childhood and 20 typically developing age-matched controls. On a simple tracking task, participants with DCD were slower and more variable than controls. The group differences dissipated, however, when the display highlighted the directional changes necessary within the next 500 ms. When the latter condition was modified to also include the full layout of the course, however, the performance of the DCD group once again decreased. This result could not be attributed to a simple distraction effect. The results suggest that distinct neural mechanisms are associated with the processing of fast visual information for online control and longer-term action preparation based on spatial layout. In skilled action, cerebellar and parietal areas process information effectively and their outputs are integrated into one smooth movement. Because the DCD group showed difficulties in steering when both types of information were present, it is likely that this integration is suboptimal.

An fMRI study of parietal cortex involvement in the visual guidance of locomotion.

Locomoting through the environment typically involves anticipating impending changes in heading trajectory in addition to maintaining the current direction of travel. We explored the neural systems involved in the "far road" and "near road" mechanisms proposed by Land and Horwood (1995) using simulated forward or backward travel where participants were required to gauge their current direction of travel (rather than directly control it). During forward egomotion, the distant road edges provided future path information, which participants used to improve their heading judgments. During backward egomotion, the road edges did not enhance performance because they no longer provided prospective information. This behavioral dissociation was reflected at the neural level, where only simulated forward travel increased activation in a region of the superior parietal lobe and the medial intraparietal sulcus. Providing only near road information during a forward heading judgment task resulted in activation in the motion complex. We propose a complementary role for the posterior parietal cortex and motion complex in detecting future path information and maintaining current lane positioning, respectively.

Stepping over obstacles: attention demands and aging.

Older adults have been shown to trip on obstacles despite taking precautions to step carefully. It has been demonstrated in dual-task walking that age-related decline in cognitive and attentional mechanisms can compromise postural management. This is yet to be substantiated during obstacle negotiation when walking. Forty-six healthy volunteers (aged 20-79 years) stepped over obstacles in their path whilst walking and performing a verbal fluency task. Using 3D kinematic analysis we compared obstacle crossing during single-task (obstacle crossing only) and dual-task (obstacle crossing with verbal task) conditions. We grouped the participants into three age groups and examined age-related changes to cognitive interference on obstacle crossing. During dual-task trials, the 20-29 and 60-69 groups stepped closer to the obstacles prior to crossing, increased vertical toe-obstacle clearance, and had reduced gait variability. In these two groups there was a small dual-task decrease in verbal output. The 70-79 group applied similar dual-task stepping strategies during pre-crossing. However, during crossing they showed reduced vertical toe-to-obstacle clearance and increased variability of obstacle-to-heel distance. Additionally, this group did not show any significant change to verbal output across trials. These results suggest that with advanced age, increased cognitive demands are more likely to have a detrimental impact on motor performance, leading to compromised safety margins and increased variability in foot placement. We conclude that younger adults utilise a posture-preserving strategy during complex tasks but the likelihood of this strategy being used decreases with advanced age.