| dc.description.abstract | Landscape services in the United States is an expansive industry, producing upwards of
$221.89B in economic output over more than 100,000 landscape companies (Hall et al., 2020).
Despite this economic strength, the industry faces growing operational challenges, particularly
related to labor shortages. Autonomous mowers offer a variety of potential benefits to the
landscape industry, such as reduced mowing times, reallocation of labor, and higher operating
efficiencies. In this experiment, four different types of zero-turn, commercial mowers were
tested in this experiment: battery-powered manual, gas-powered manual, battery-powered
autonomous, and gas-powered autonomous. Testing was conducted on two landscape designs—a
simple square plot with no obstacles and a complex plot with obstacles—to analyze variations in
efficiency. Turfgrass species had no statistically significant effect on energy consumption for
either mower type. Across tests, 1.23 mL of gasoline corresponded to 1 Wh of battery energy.
Autonomous mowers took much longer to complete the free plots compared to the manual
mowers, increasing average time to cut to 33.7 minutes per acre. These data suggest that the
most suitable context for these mowers is large, unobstructed turfgrass areas, such as sports
fields, municipal parks, and large campus grounds, where path complexity and obstacle density
are minimal. However, as autonomous mowing equipment continues gain popularity, many
consumers have concerns and hesitations regarding the safety of autonomous mowers. Two
different types of technologies were tested in this experiment: one autonomous mower equipped
with 4 stereo RGB cameras, and one autonomous mower equipped with 5 radars. Both types of
detection technologies were evaluated with different obstacle sizes, colors, and approach
angles. Overall, the Wright mower equipped with stereo RGB cameras was more effective at
detecting all obstacles, resulting in a 93.8% probability rate, compared to the Bobcat mower
equipped with a radar system, which had a probability of 61.8%. These data suggest that large
platform autonomous mowers cannot be unsupervised in an area where adult and child obstacles
may be present while the mower is working. | en_US |
