ThrowBot: Design Considerations for a
Man-Portable Throwable Robot
Mitch Barnes,*a H.R. Everett**a
Pavlo Rudakevych b
a SPAWAR Systems Center, Code 2371, 53560 Hull Street, San Diego, CA 92152-5001;
b iRobot West, 3485 Sacramento Dr. Ste. C, San Luis Obispo, CA 93401
ABSTRACT
无锡英语翻译The pocket-sized ThrowBot is a sub-kilogram-class robot that provides short-range remote eyes and ears for urban combat. This paper provides an overview of lessons learned from experience, testing, and evaluation of the iRobot ThrowBot developed under the Defen Advanced Rearch Projects Agency (DARPA) Tactical Mobile Robots (TMR) program. Emphasis has been placed on investigating requirements for the next generation of ThrowBots to be developed by iRobot Corporation and SPAWAR Systems Center, San Diego (SSC San Diego) Unmanned Systems Branch. Details on recent
evaluation activities performed at the Military Operations in Urban Terrain (MOUT) test site at Fort Benning, GA, are included, along with insights obtained throughout the development of the ThrowBot since its inception in 1999 as part of the TMR program.
Keywords:robotics, ThrowBot, man-portable, urban combat, unmanned ground vehicle.
1.0 BACKGROUND
Although the development of unmanned-systems technologies has significantly advanced over the years, the actual u of robotics in the military has remained elusively just over the horizon; that is until the recent deployment of man-portable systems in theatre for explosive ordnance disposal (EOD) tasks. For the first time in history, the warfighter is able to provide feedback on operation and ufulness of various man-portable systems in combat and assist in directing the future development of robotic platforms as a warfighting tool. Tradeoffs with regard to state-of-the-art technology integration versus size, weight, speed, agility, and mobility have helped minimize the drawbacks and maximize the benefits to the soldier. This spiral development process has identified a need for a throwable sub-kilogram rolling-camera robotic capability, or ThrowBot, that allows for remote reconnaissance of unknown or hostile areas.
Initial prototypes for the ThrowBot were developed under the TMR program, sponsored by DARPA in the late 1990’s, and transitioned to SSC San Diego at the end of FY-02. This provided SSC San Diego a varied mix of prototypes to test requirements for future throwable platforms. SSC San Diego, funded by the Office of the Secretary of Defen (OSD), was tasked to take laboratory prototypes developed under the TMR program and determine a path forward to develop a fieldable throwable robotic system.roms
1.1 Concept of Operations
At ba or in a vehicle (HMMWV), the soldier ts up the ThrowBot and
its associated controller to charge up the batteries. While prepping for a
mission, the ThrowBot and controller are removed from the charging
stations and stowed either in a backpack or cargo pockets.2010年12月六级听力
During a combat mission, the system remains stowed until there is a
pau in operations, and it is desired to know if hostiles are in a particular
* mitchell.barnes@navy.mil; robot.spawar.navy.mil/
** everett@spawar.navy.mil Figure 1. ThrowBot concept.
nearby vicinity. The ThrowBot is removed from the carrying compartment, activated, and then tosd down a corridor, up a stairwell or into a window as shown in Figure 1. The platform can be thrown by a single person or launched into an upper window or rooftop using an improvid slingshot consisting of surgical tubing held by two soldiers and operated by a third. Landing on a top
floor, the remotely operated platform will be able to bypass typical obstacles and travel down stairs. The operator can evaluate the resultant video before determining the next cour of action. The ThrowBot can also be ud to arch for and asss booby traps, enemy personnel, and improvid explosive devices (IEDs).
1.2 Initial Requirements
From previous experience in developing a variety of robotic systems and discussions with prospective urs, an initial t of requirements for the ThrowBot-class platforms has emerged. In terms of vision, the camera must be low-light/night capable while not compromising its ability to look into bright sunlit areas without any washout of the picture. It should be possible to point the camera in any direction and elevation angle, zoom the camera on an item of interest and still have a wide field of view for driving in a cluttered environment. The associated video display should also be viewable in direct sunlight.
Effective communications range is also a key factor in order to provide the warfighter sufficient standoff from the enemy while operating the platform in an open area, with a non-line-of-sight capability through building walls. The platform needs to be able to operate in a cluttered environment and should not get hung up in debris such as clothing, rocks or wires.
Deployment in dert scenarios typically encountered in Iraq and Afghanistan impos a requirement to operate in thermal conditions of 100-140 degrees Fahrenheit.
fuckyou什么意思2.0 TECHNOLOGY ASSESSMENT
2.1 Early Prototypes
In the early TMR ThrowBot efforts, many mobility methods were considered, but not all were brought to the prototype level, and conquently many dynamic methods of overcoming obstacles have not been fully explored. This ction provides an asssment of current and past efforts in throwable platforms, including the three prototypes that were lected for further testing by SSC San Diego.
2.1.1
Two-Wheeled Cylinder
Figure 2. Two-wheeled cylinder. One of three prototypes developed by iRobot Corporation for the 1999 subcontract to
Draper under the TMR project was the two-wheeled ThrowBot. With dimensions similar
to a 12-ounce soda can, it posss a cylindrical body with a large wheel on each end and a
spring-loaded tail. This design potentially allows for large-diameter optics to lookgets
sideways through the wheel hubs. However, while the design can lf-right, its ability to
negotiate terrain or climb over small obstacles is extremely limited, and it has a strong
tendency to yaw back and forth while moving, riously degrading the video. A number of
other TMR participants pursued similar two-wheel approaches, but all were eventually abandoned for the reasons cited.
2.1.2 Six-Wheeled Brick
The solid brick-like six-wheeled platform posss a rectangular body, with three wheels
on each side geared together for full six-wheel drive with skid steering. A paddle, which
stows on the top, is ud to right the vehicle if it becomes inverted, and to help push the
vehicle over obstacles. The vehicle can climb over obstacles as tall as its wheelba, such
as most street curbs. Figure 3. Six-wheeled Brick.
cat翻译2.1.3 Four-Wheeled
Brick
泰坦尼克号2杰克归来
Similar in body construction to the six-wheeled Brick, the four-wheeled platform was
modified by DRAPER laboratory to have larger wheels that make it able to climb bigger
obstacles and a slower drive speed to allow greater control by the ur. This design and its
six-wheeled variant above are the most viable candidates built to date.
Figure 4. Four-wheeled Brick.
Weight-Shifting Sphere
2.1.4 Inflating
A compact package when stowed, this design inflates into a spherical ball after being
所以的英文thrown or emplaced. Limited mobility was achieved by weight shifting the internals, and
the inflating skin was susceptible to sharp objects in the environment. The platform also
suffered from nsitivity to wind, and therefore was not pursued further.
Figure 5. Inflating sphere.
2.1.5 SpinyBall
Developed by DRAPER, the SpinyBall starts off as a softball sized sphere making it highly
portable and easily deployed. When deployed the operator can extend the spiny wheels,
exposing the camera and giving the platform a unique solution to driving over difficult
terrain. This solution has fairly good mobility but is a complex design with many moving
parts, and therefore expensive to build and harder to maintain.
Figure 6. SpinyBall.
2.1.6 Rebound with Flippers
iRobot specifically the Rebound for testing the dual-flipper concept. The Rebound
ThrowBot is a quickly constructed prototype designed around a commercial toy car with
dual external flippers mounted on the rear. In the laboratory, it has demonstrated superior
terrain transverability and camera placement but suffered from poor reliability and lacked
rvo control on the flippers.
Figure 7. Rebound.
2.2Evaluation Platforms
The six-wheel Brick, four-wheel Brick, and Rebound with flippers were the clost to meeting the stated requirements and thus chon by SSC San Diego for further evaluation. The initial prototypes were intended as mobility demonstration platforms, allowing urs to experience the current state of technology, and engineers to investigate where improvements needed to be made to produce fieldable units. Although the prototype concepts were never intended for durability testing, the two Brick options are hardened to withstand casual impacts, and the design allows for further environmental ruggedization.
3.0 USER EVALUATION
In late May 2004 SSC San Diego took the two Brick ThrowBot prototypes and the Rebound ThrowBot shown in Figure 8 to the Fort Benning McKenna MOUT site, where a simulated town rests in the middle of a Georgia forest, complete with furnished buildings, city streets and staffed by army personnel.
Figure 8. Two Brick ThrowBots (foreground center) and one dual-flipper Rebound ThrowBot.
The three ThrowBot platforms were prented to urs in teams of four to six per group. After the teams were briefed on the intent of the ThrowBot evaluations, each ur was provided the opportunity to drive the platforms through an obstacle cour in direct view. Next, each ur was given an opportunity to drive the platforms using only video feedback from the operator control unit (OCU). Once each ur had mastered the prototypes, the robots were dropped into a realistically cluttered urban environment, and the teams were requested to map the rooms and identify hidden inert weapons, such as grenade launchers, AK-47s, and trip-wire mines (Figure 9).
Figure 9. Weapons (inert) hidden at Fort Benning (left) and a map of weapons drawn by urs during evaluation (right).
4.0 TECHNICAL
ASSESSMENT
Many upgrades for the ThrowBot were identified as a result of interacting with each platform, as well as by direct obrvation of operating personnel and surveys distributed to urs (e Appendix A and Appendix B). The upgrades range from simple ur-interface concerns to mission-critical issues, such as communications range and portability.英语六级题型
4.1 Mobility
Urs reported that the six-wheeled ThrowBot felt unnatural to remotely drive due to the low height of the camera and its high speed. They reported “jitteriness” causing overcorrection while turning the six-wheeled platform. While the four-wheeled ThrowBot felt more natural to most urs due to the higher camera mounting, it was slower than some would have preferred. The speed tting for the individual ThrowBots is a variable parameter bad on the gear ratio and ttings in the OCU. This issue could be easily addresd in any future design by a readjustment of the parameters to find the ideal ratio of responsiveness and control. Steering overcorrection can be addresd by replacing the open-loop speed control on the prototypes with clod-loop control. Overall the mobility of the platforms was adequate to perform most missions and had the ability to overcome many common obstacles, as en in Figure 10.
Figure 10: Six-wheel ThrowBot climbing over open grate.
4.2 Portability
Since portability is a major concern for an item intended to be carried by the warfighter in theatre, weight was kept to a minimum during the design of the ThrowBot. However, due to its awkward shape, it was not comfortable to walk around with the platform in a standard cargo pocket, as en in Figure 11.
Figure 11. Pha I ThrowBot was designed to fit into standard thigh pocket.
To make transporting the ThrowBot easier, veral urs suggested incorporation of a customized pouch that could be clipped to a standard-issue backpack. Another suggestion was to design detachable wheels that could be taken off easily without tools, making the system more portable. This would allow for future development of specialized wheels for different environments and also facilitate modular maintenance.
4.3 Low-Light
Visibility
After entering a few buildings, it quickly became clear that operation indoors and under furniture brought the ThrowBot into conditions of extremely low light. Some areas of the environment could not be fully perceived by the teams due to the lack of visibility under the conditions. Two possibilities are being evaluated for a future design: 1) to equip the unit with a more expensive low-lux camera that can e in almost any lighting condition, 2) to install low-cost lighting on the ThrowBots to illuminate the camera field of view. While lower in cost, the latter option introduces additional battery drain and increas the chance of detection during covert operations.
幸福 英文Life
4.4 Battery
The general connsus was that the 2-hour battery life was more than adequate for typical tasks. In the future, built-in rechargeable batteries would simplify logistics support in-theater, but replaceable alkaline batteries allow the unit to return to rvice immediately.
4.5 Communications
Range
Communications range was experimentally found to be approximately 40 meters of open area or through two cinderblock walls of a building, which the majority of urs claimed was sufficient for many missions. A more
