UC Davis

Farmers need more than a single trusty tractor to handle most of the jobs and tasks on their property. Tractors are useful for tasks that require high power such as seeding and plowing. However, for some daily-farm tasks that need lower power, tractors are being replaced with All-Terrain Vehicles (ATVs). While ATVs cannot match up the strength of tractors, their versatility and relatively lower costs allow farmers to efficiently perform some of the jobs and tasks in a farm that do not require high power, such as transporting supplies, mowing grass, checking fences, herding calves, and carrying firewood. However, the increasing use of ATVs as a utility vehicle adds a heavy burden to the American public health system. The U.S. Consumer Product Safety Commission (CPSC) estimates that in the past ten years, ATV-related incidents have resulted in over 6,500 deaths and 925,000 hospitalizations. In addition, the annual cost of lives and health care from ATV-related incidents has increased almost five times in the past decade, reaching $22 billion dollars spent in 2016. ATVs have engineering features such as low-pressure tires, narrow wheelbase, and high center of gravity that make them prone to rollover when riding on rough and uneven terrains, or steep slopes (all common scenarios in farms and ranches). Indeed, agriculture is the major contributor for incidents involving ATVs, accounting for 50% and 65% of all occupational-related injuries and deaths, respectively.Based on data from the CPSC, youth younger than 16 years old are the leading victims of ATV incidents. Furthermore, a previous study found that ATVs are one of the primary sources of vehicle injury for youth on farms, causing 63% of the vehicle-related injuries. Moreover, according to “AgInjuryNews.org” in 2021 the number of reported fatalities and nonfatal injuries in the U.S. among youth caused by ATV were 52 and 26, respectively. Data from the 2019 National Electronic Injury Surveillance System revealed that youth younger than 18 accounted for 36.8% of all ATV-related injuries. Over 15% of those injuries occurred on farms or ranches. Furthermore, several studies identified a correlation between ATV-related injuries of children and their readiness to ride, including their strength and anthropometry, among other characteristics. For these reasons, two studies of this dissertation focused on evaluating the capabilities and limitations of youth operating utility ATVs. Furthermore, previous studies showed that crash location is an important risk factor for ATV-related incidents. Most ATV crashes on farms and ranches result in traumatic injuries where the rider needs immediate care but is unable to seek help because they are severely injured. Further compounding the issue, most of these crashes occur in isolated areas of hard access and without reliable and constant cellular service. Thus, making it challenging to contact emergency medical services (EMS) promptly and receive first-aid care in a timely manner. Therefore, one study in this dissertation aimed at developing and testing a low-cost, ATV crash-prediction-and-detection device (AgroGuardian) that immediately alerts EMS, even when the rider is unable to do so, and there is no cellular service available. In the first study, potential discrepancies between the required activation forces of eight controls of fifty-four utility ATVs and the strength of youth of varying ages (6-20 years old), genders (males and females), and strength percentiles (5th, 50th, and 95th) were evaluated. In addition, we also assessed if youth strength is enough to push the ATV off if they are pinned underneath it, which is a common post-rollover scenario that can result in death by mechanical asphyxia. A handheld force gauge, a button load cell, and a pressure glove were used to measure the activation forces of the main controls (handbrake, footbrake, handlebar, throttle lever, ignition switch, headlight switch, hand gearshift, and foot gearshift) of the utility ATVs. The activation forces of the ATVs’ controls were compared with the corresponding strength of youth found in previous reports. The results of this first study demonstrated a physical mismatch between the forces required to operate ATV controls and youth’s strength. Turning the handlebar, pressing the footbrakes, and pushing the ATV off were the most difficult tasks for ATV operation. For instance, youth aged 6-10 would be able to activate the footbrake of only 64% of the evaluated ATVs. The inability to depress the footbrake affects the youth’s ability to reduce the speed or stop the ATV and can prevent the operator from diverting from obstacles or potential bystanders. The results were even more striking when considering the ability of a youth operator to push the ATV off if pinned underneath it. Less than 13% of all evaluated vehicles could be pushed off by male operators aged 16-20 years old of the 95th strength percentile, the strongest subjects of this study. These discrepancies compromise the youth’s ability to ride ATVs, increasing their risk of crashes. In the second study, potential inconsistencies between the operational requirements of utility ATVs and the anthropometric dimensions of youth were evaluated through virtual simulations in an ergometric software (SAMMIE CAD). The simulations were performed to assess eleven reach-based ATV fit guidelines proposed by several ATV safety advocacy organizations (National 4-H council, CPSC, IPCH, and FReSH). In total, seventeen utility ATVs along with male-and-female-youth of nine different ages (8 – 16 years old) and three height percentiles (5th, 50th, and 95th) were evaluated. The results demonstrated a physical mismatch between ATVs’ operational requirements and youth’s anthropometry. For example, male-youth aged 16 years old of the 95th height percentile failed to pass at least one out of the eleven fit guidelines for 35% of all vehicles evaluated. The results were even more concerning for female operators. Female youth 10 years old and younger (from all height percentiles) failed to pass at least one fit guideline for all (100%) ATVs evaluated. As such, youth should not ride utility ATVs. The results from these first two studies provide quantitative and systematic evidence to modify/update current ATV safety guidelines. Furthermore, youth occupational health professionals could use the present findings to prevent ATV-related incidents in agricultural and other settings. Lastly, in the third study, a device (AgroGuardian) was developed to make online predictions of the likelihood of an ATV rollover based on the ATV characteristics (e.g., length, width, height) and riding conditions (e.g., speed and attitude) and alert the rider in real-time if this likelihood is above a pre-set threshold. In addition, AgroGuardian automatically notifies EMS and emergency contact(s) when a rollover is detected even though no cellular service is available, and the rider is unable to take action. AgroGuardian includes an embedded data logging system, a smartphone application, and a remote database. The embedded system includes a 3-axis Inertial Measurement Unit (IMU) for attitude (roll, pitch, and yaw) estimation, a low-cost Global Positioning System (GPS) for position estimation, and a Rock7 modem for off-board communication. To reduce the system’s attitude error, a Madgwick filter was implemented to fuse the data from the sensors of the IMU (accelerometer, gyroscope, and magnetometer). Similarly, GPS and IMU data were fused through an Unscented Kalman Filter (UKF) to improve the ATV’s position estimate. The smartphone application was developed to receive inputs from the users regarding their machines (e.g., make, model, and characteristics such as width and length), to log information about emergency contacts, and to allow them to interact with their ATV data. The ATV’s riding data collected by the sensors in the embedded system along with the ATV characteristics inputted from the user via the smartphone app are fed to a deep neural network to make online rollover predictions. An emergency signal along with the ATV’s coordinates are sent off-board through the Rock7 modem and received in the remote database when a rollover is detected by the system. This emergency signal is then processed and sent to EMS and emergency contact(s). The performance of the proposed device was assessed through experimental tests simulating rollover incidents and normal riding conditions. The results indicate that: (1) the device has a rollover prediction system with an accuracy superior to 99%; (2) the device has a rollover detection system with an accuracy superior to 99%; (3) the device has a fast EMS notification time (40.70 s); (4) ATV localization presented an accuracy of 2 m.