Transcript poster template - Computer Science & Engineering

Robot Highway Safety Markers
Department of
Mechanical Engineering
Dr. Shane Farritor
University of Nebraska, Lincoln
Sponsored by The National Academy of Sciences IDEA Program
and The National Science Foundation
Computer Science and
Engineering Department
Dr. Steve Goddard
Objective
Real-time Systems Research
Proper traffic control is critical in highway work zone safety.
Traffic control devices such as signs, barricades, cones, and plastic
safety barrels are often used. Accidents can occur because of
improper work zone design, improper work zone housekeeping,
and driver negligence. Automated safety devices could improve
work zone design and housekeeping and therefore increase safety.
The control operations of the barrel robot demand both logical and temporal correctness. Temporal correctness requires
these operations to be computed and executed within allowable response times. Instead of a single control loop for
these operations, real-time task sets are used. In real-time scheduling theory, a task is a schedulable unit of work that
executes with a known pattern. The real-time task set provided by microC OS-II, a real-time operating system driving
the barrel robot, is used to implement the local planning and control operations. The use of real-time task sets allows
temporal correctness to be analyzed and guaranteed in advance. It provides maintainability and extensibility; if the
processor is upgraded, only the tick-rate of the processor needs to be updated in the software. It reduces latency
between the receipt of a waypoint from the lead robot and the time at which the barrel robot begins to move. As the
robot is moving to a waypoint, a new waypoint is calculated.
The robot highway safety marker serves as a suitable application to study research problems in real-time systems. For
example, the limited power in the barrel robot requires the system to be aware of energy constraints. An interesting
area of research is scheduling algorithms to maximize the battery life in such an embedded system. Dynamic voltage
scaling (DVS) algorithms aim at reducing the energy consumption of the system by operating the CPU at a lower
frequency and thus operating at a lower voltage. An ongoing research project is evaluating a proposed DVS algorithm
that works in conjunction with the earliest-deadline-first (EDF), dynamic priority scheduling algorithm. The
algorithm focuses on the sporadic task model, which
puts only a lower bound on the time separation
interval between the release of jobs of the same task.
This flexibility of the sporadic model makes it
conducive to many applications. The proposed
algorithm guarantees that each job meets its deadline
while saving the maximum amount of energy.
Description
Safety barrels guide traffic and serve as a visible barrier between
traffic and work crews. These barrels consist of a brightly colored
plastic drum (approximately 130cm high and 50cm in diameter)
that is attached to a heavy base. Often, hundreds of barrels are
manually placed in a typical work zone. The Robotic Safety Marker (RSM) replaces the heavy
base of a typical safety barrel with a mobile robot. The mobile robot can transport the safety barrel
and robots can work in teams to provide traffic control. Shown above is the robot base next to a
plastic safety barrel. The RSM can self-deploy and self-retrieve, removing workers from this
dangerous task. The robots can move independently so they can be deployed in parallel and can
quickly reconfigure as the work zone changes. The robot base has two electric motors. Each wheel
is driven by its own motor. A castor supports the rear of the robot and the reversible, variable
speed motors allow the robot to move in any direction as well as turn in place.
Planning and Control
Although, each robot moves individually, a single lead robot (general) provides global planning
and control and issues commands to each barrel robot (troops). The lead robot plans the path,
communicates waypoints, and monitors performance of each barrel robot. The graph below (left)
shows the global plan for two robots to move from random locations on the roadside shoulder to
positions in a taper, to close a lane. To obtain the desired robot path, a parabola is created between
the initial position,and orientation, and the final position. Each barrel robot does not have
knowledge of other robots, and performs
only local tasks. It receives a waypoint
from the lead robot and creates
localized
positions between
the initial position
and the desired
waypoint.
By: Xiangrong Shen, Jason Dumpert,
Field Tests
The RSM has been tested in field environments. The
picture on the right shows all robots operating as a
team to close the right lane of a two-lane road. The
desired and actual paths taken in
the tests were plotted as shown in the figure below (lower right). The
symbols on each actual path represent where individual robots
reached a waypoint and their position was updated.
Lines connecting symbols only approximate robot motion
between points. The maximum deviation from the desired
path was 23 cm and the maximum final error for all robots
was 11 cm. This accuracy is well within the requirements
for barrel placement and exceeds the accuracy of current
human deployment.
Chon-Ming Lee, Rohini Krishnapura, Ala Qadi