SAMMIE CAD provides a 3-D environment and full control of human mockups, which makes it possible to evaluate those complex interactions. The simulations performed in SAMMIECAD consisted of: creating 3-D human mockups; creating 3-D ATV mockups; and integrating and in the virtual environment to simulate their interaction. For each simulation, the correct reach posture was achieved by positioning the human limbs according to the specific task’s requirement. For example, a seated position was adopted when evaluating fit criterion 10 , as shown in Fig. 3a. On the other hand, a standing straddling posture was selected when evaluating fit criterion 4 , as shown in Fig. 3b. Some criteria involve the youth reaching a specific control . The feature ‘‘Reach” under the ‘‘Human” menu on SAMMIE CAD was used to evaluate the ability of the youth mockups to reach the selected controls. The ‘‘Reach” was set as ‘‘Absolute,” and ‘‘Object Point” was set as ‘‘Control.” When the selected control could be successfully reached, the software would display an animation of the human limb reaching the desired object . On the other hand, if the control was out of reach, SAMMIE CAD would show an error window and display the required distance for the human limb to reach the desired control . Simulations involving buttons and levers were performed with the fingertip of the index finger or the thumb, accordingly. All controls on the right side of the ATV were simulated with the right hand/foot, and all controls on the left side of the ATV were simulated with the left hand/foot.
Specific controls that required using both hands, such as the handlebars, seedling starter trays were simulated with both hands. Criteria 1, 2, 3, and 11 were evaluated through Matlab because their assessment required the computation of simpler calculations, such as the distance between the rider’s knee and the ATV’s handlebars. Matlab also provided the ability to automate the calculations for a more efficient data analysis. A code was generated based on conditional statements to assess whether riders’ anthropometric measures conformed to the constraints imposed by the ATV design. For instance, when evaluating criterion 1, the distance between the ATV footrests and the handlebars minus the rider’s knee height must be greater than 200 mm . For each reach criterion, riders received a binary score . Riders with a total score of 11 were classified as ‘‘capable of riding the ATV.” On the other hand, riders with a total score below 11 were classified as ‘‘not capable of riding the ATV.”In order to validate the results of the virtual simulations, an experiment including three adults and one study ATV was carried out. Each subject had completed an ATV safety riding course prior to the experiment and was awarded a certificate from the ATV Safety Institute . The capability of the subjects to fulfill each fit criterion was evaluated and recorded. For the field tests, a measuring tape graduated in mm was used to measure distances and a digital angle finder to measure angles. To assist in some of the angle measurements, a straight edge 4800 ruler and a mag-netic level were used. The anthropometric measures of the subjects were taken with a body-measuring tape and then used as input in SAMMIE CAD to create 3-D mockups.
The results observed in the experimental setting were then compared to those observed in the virtual simulations through the Cohen’s Kappa coefficient , which is a statistic widely used to measure inter-rater reliability for qualitative items . A Z-test was performed to evaluate whether the value of K was statistically different than zero, which would imply that the virtual simulations are reasonable.Seventeen ATV models were evaluated from eight different manufacturers. Engine capacity ranged from 174-686 cc, with most vehicles in 100–400 cc . Moreover, 58 % of the ATVs evaluated included electric power steering , 4 wheel-drive , solid suspension , and manual transmission . Findings of individual reach criteria for the ATV models are presented in Tables 2 and 3, for males and females, respectively. The last column of those tables represents the percent of observations for which riders scored 11 points . Criterion 1 seemed difficult for 16-year-old-males of the 95th body-size percentile. This result may be attributed to the height of these subjects, which decreases the gap between their knee and the handlebars .Unlike criterion 1, criterion 2 did not present any difficulty for the virtual youth . Indeed, virtual subjects of all ages, body-size percentiles, and genders succeeded in this criterion for all evaluated vehicles. Criteria 3, 4, 6, 7, 8, 9, 10, and 11 all presented a similar trend where young riders do not conform well to these criteria, but older riders do . The contrast in success rate among subjects of different ages and height percentiles are likely also attributed to the variations in height among the subjects. For example, virtual 8-year-old-female riders of the 95th percentile did not pass criterion 5 for any of the evaluated ATVs. In contrast, their 16-yearold-counterpart passed the same criterion for 75 % of the evaluated ATVs , a surprising difference of 75 %. The results from Tables 2 and 3 indicate that 8-year-old youth would probably not be able to control utility vehicles when traversing rough or uneven terrains . This finding likely explains the fact that youth are more subject to loss of control events than adults .
The results of the simulations related to Criterion 7 indicated that youth 9 years old and younger are more likely to lean forward over 30 when raised off the seat to reach the handlebars of agricultural ATVs. As a result, the center of gravity of the ATV can shift forward, thus increasing the chances of a tip over. Lastly, some results of the simulations related to Criterion 5 were concerning. Males up to 11 years old and females up to 13 of the 50th percentile passed this criterion for less than 50 % of the evaluated ATVs.The percent of ATVs in which riders passed all criteria is presented in Fig. 4. The main finding is that certain youth should not ride most utility ATVs. For instance, the average male operator aged 16 passed all 11 safety criteria for less than 60 % of the evaluated vehicles. That number decreases sharply for younger youth or youth of the same age but smaller height percentile. A similar trend was also observed for female operators.The results of the validation tests are presented in Table 4 and summarized in a confusion matrix . In the confusion matrix, the outcome of the test is labeled in both horizontal and vertical axes. The horizontal axis represents the number of outcomes predicted by the virtual simulations, and the vertical axis represents the ground truth data . The results of the virtual simulations were very close to those of the field tests, with a total accuracy of 88 %. The Z-test determined that the Cohen’s Kappa coefficient was significantly greater than zero , botanicare trays indicating that the virtual simulations are reasonable. This approach to evaluate ergonomic inconsistencies between youth’s anthropometry and the operational requirements of ATVs proved to be an effective and accurate technique. Not all results of the virtual simulations matched those of the field tests. One unexpected result is related to criterion 6 . It was observed that the mean angle between the riders’ upper leg and the horizontal plane was 16.7 , slightly above the recommended threshold . Similarly, two subjects failed to pass criterion 5 in the actual field tests but passed it in the virtual simulation. During the field tests, riders were asked to sit comfortably as if they were just about to start riding the ATV. We argue that it would be possible for riders to adjust their way of sitting so they would pass both fit criteria; however, it would not result in the most ergonomic posture from the rider’s standpoint. On the other hand, in the virtual simulations, our ultimate goal was to place the 3-D subjects’ mockups to physically conform to the proposed fit criteria. Thus, it was impossible to predict whether the final adopted postures in the simulations would match those selected by the riders in the validation tests. Therefore, we argue that despite some outcomes of the virtual simulations did not match those of the field tests, the results of the virtual simulations are still reasonable. One just has to be cognizant that the outcomes of the virtual simulations represent a hypothetical scenario where the rider is able to attain a posture based on their anthropometric measures relative to the ATV, not on their preferences.
This study evaluated limitations in youth’s anthropometric dimensions when riding commonly used ATVs. Using a combination of actual field measurements and a novel digital simulation approach, the present study evaluated 11 ATV fit criteria for youth. The major finding was that youth are not recommended to ride adult-sized ATV models, which is a common practice in the United States , 2010; Jennissen et al., 2014. This finding raises serious concern regarding youth’s ability to ride ATVs, especially when unsupervised.The present findings outlined that some youth are too small, which makes them incapable of properly reaching the vehicle’s hand/foot brakes, resting their feet on the footrests, or having to lean forward beyond 30 to reach the handlebars when rising off the seat. Failing to activate the ATV brakes limits the youth’s ability to reduce the speed or to stop the vehicle, which likely prevents them from avoiding unexpected hazards, such as obstacles or bystanders . In fact, previous research has shown that a significant number of ATV incidents include hitting a stationary object . In addition, the inability to place the feet on the footrests when not breaking the ATV entails a functional loss of control of the vehicle. ATV LCEs occur frequently and are a significant cause of injury and death in agriculture . This finding indicates an opportunity for manufacturers to consider changing the design of their machines, allowing riders to adjust the ATV’s seat height, which would likely reduce longitudinal torso impact while traversing rough and uneven terrains. Furthermore, leaning beyond 30 can cause the ATV to tip forward, resulting in a rollover. Most ATV-related crashes on farms and ranches, especially those resulting in deaths, involve rollovers . On the other hand, some youth are too tall, which decreases the clearance zone between their legs and the handlebars. A clearance zone smaller than 200 mm makes it difficult for the rider to properly reach and steer the handlebars . Consequently, riders may lose control of the vehicle or have difficulty keeping it at a safe speed. As mentioned before, these series of events can lead to injuries and deaths.Furthermore, despite some results showing that youth are capable of riding many of the ATVs evaluated in this study, other risk factors such as experience, psychological, and cognitive development cannot be overlooked . Youth who are high in thrill-seeking are more likely to engage in risky ATV riding behaviors, regardless of their safety awareness . Those cases require external interventions, such as changes in legislation, improved ATV design, and use of crush protection devices .The results of this validation experiment showed that some riders failed criteria 5 and 6 even though they seemed able to operate the study vehicle comfortably and safely according to our ATV safety research team. Particularly, subjects 1, 2 and 3 presented elbow angles of 129 , 170 and 172.5 , respectively. While fit criterion 5 recommends an elbow angle between 90 and 135 , it is not uncommon to see motorcycle riders reporting comfortable elbow angle values up to 168 . Moreover, subjects 1, 2, and 3 presented upper leg angles of 14 , 14.7 , and 21.4 , respectively . A previous survey regarding motorcycle riders’ perceived comfortable posture reported optimum upper leg angles as high as 23 . It is our understanding that fit guidelines 5 and 6 are rather conservative, and their proposed thresholds may rule out riders that are perfectly able to ride utility ATVs safely and comfortably. As such, we propose some modifications to those fit guidelines. First, we recommend that the rider’s elbow angle should be between 90 and 170 as long as the rider feels comfortable steering the handlebars and is able to pass fit criteria 8 and 11 .