Engr 102 Project Design Faculty of Drexel University

ENGR 102

Freshman Engineering Design

PROPOSAL

AND THE

ENGR 102 PROJECT DESIGN FACULTY OF DREXEL UNIVERSITY

TEAM MEMBERS

Submitted in partial fulfillment of the requirements for

Freshman Engineering Design, ENGR 102, Design Project

Submitted on ______3/12/07______

{date}

- 1 -

Lego Mindstorm NXT Rescue Robot

1. PROBLEM STATEMENT

In disasters, human rescue workers are usually put in much danger. There are dangerous situations, which involve many obstacles. Besides the obvious risks of physical injury and death, many rescuers have to deal with psychological effects after the disaster. If a robotic system could be designed that minimizes human presence at the disaster site, fewer rescuers would be put in immediate physical and psychological danger (Disaster Rescue and Response Workers 1). We will design a Lego Mindstorm robot to simulate the extraction of a casualty autonomously.

1. INTRODUCTION

In many ways, search and rescue robotics is still in its early stages. Many professors and professionals admit that robots can be slow and awkward at times, especially while some robots may still be in the prototype stage of development (New Scientist 1). Some times, robots are ill used. For example, a robot that usually worked for the bomb squad was commandeered to rescue trapped miners. The robot was programmed to dig, but eventually the robot malfunctioned due to the fact that the robot was not designed for that particular job (New Scientist 3). However, the biggest complaint in the world of robotics is the lack of monetary funds to continue research and development for search and rescue. Many researchers believe that because search and rescue does not necessarily create profit, companies and the government often overlook it (New Scientist 2).

Despite this current situation, many new solutions have been developed for search and rescue. Now, engineers are developing new technology and utilizing existing technology to help save lives. Many designs like DARPA’s Polybot use animals like snakes to create the design: the Polybot wiggles side to side in order to squeeze through small spaces in mine disasters. The robot asterisk has six legs with joints which correspond to spiders and insects (Money Not Art 2). These robots can cling to ceilings to detect survivors when passages are obstructed or blocked. Other designs like the micro-rover use wheels to navigate and sensors to detect the presence of an object (Boyd 1). This allows rescuers to detect survivors with minimal risk to rescue workers. Also, these robots are all equipped with microphones to pick up sounds from trapped victims, and thermal cameras to detect moisture and body heat. Micro-rovers can enter into a collapsed mine and use laser rangefinders to create a detailed 3-D map of the disaster area. This enables rescue workers to locate and formulate a plan to overcome future obstacles. In the future, many more possible solutions will surface.

1. PROPOSED DESIGN
2. Criteria

For the solution to be deemed successful it must follow the thresholds that are set by the professor. However, those are just the base points to meet. We want the solution to be able to accomplish the objectives:

There are no real regulations to consider since our solution is made from LEGOS®. If we were going to sell our device it would have to be priced around at least the price of the NXT LEGO Mindstorms™ kit, which is $250. If our solution has other parts that aren’t included in the LEGO® kit, then the price of those parts would also be included. A product like this one would be marketed towards kids that are old enough to understand programming. There aren’t any safety concerns, but there are many limitations to the NXT™ kit. This kit has so many different parts that making something can take a lot of time, and needed parts come in small supply. To accomplish the thresholds and objectives, each area of the robot must be evaluated to determine the best approach to the problem. The Overall QFD will guide our decisions on importance once a design is chosen. This QFD was compiled initially from the criteria and the role each aspect of the robot plays in completing said criterion.

Table 1: Overall Quality Function Deployment (QFD)

0-5, 5 = Most Important

3.2Alternatives

3.2.1 Object Detection

One of the first major problems is how we will be able to detect the object. There are several different sensors that we can use to detect the object such as the light sensor, touch sensor, and/or the ultrasonic sensor (Pictures in Appendix C). The light sensor is able to distinguish between varying degrees of dark and light values and can be used to follow a line. The touch sensor senses pressure and knows when it “bumps” into something. The ultrasonic sensor detects what distance an object is away from it.

3.2.2 Movement

Movement is the next major problem. The different ways of movement include wheels, tracks, and/or to just keep the device from moving. Wheels provide the greatest speed and the greatest amount of adaptability. Tracks are stable but consist of a lot of pieces. An immobile structure could operate faster, but it is site specific.

3.2.3 Object Manipulation

The device that we will use to actually grab the object is the next major problem we have to overcome. There are three candidates in this category: claw; conveyor belt, or a device that can drag the object. The claw is known to work but has a lot of pieces. The conveyor belt is effective, but consists of pieces that are not in the NXT™ set. The “dragger” is very light on pieces, but is less reliable than the claw.

3.2.4 Navigation

There are three ways to navigate the simulation: by design, by sensors, and by motors/ hard-programming. Designing navigation would use a rope or string to find and return to the casualty. This rope would not be part of the NXT™ set. Sensors can be programmed to get to the casualty and back to the triage site. The motors can also be hard coded to go to the exact spot of the casualty.

3.2.5 System Interaction

All systems discussed above will be controlled by the NXT™ computer. This computer has a LCD interface for basic programming on the face of the “brick”. Maximally, three sensors and motors can attach to the computer through a six-pin connection similar to loaded telephone wire. A USB or Bluetooth connection is used to connect the NXT™ with a personal computer. Several programs can then be used to program the NXT extensively.

(Above: Picture of NXT™ computer)

1. METHOD OF SOLUTION

In order to determine the best design out of all of the different alternatives, it is necessary to experiment with the alternatives within the given boundaries. This requires that we construct a basic design of each alternative, model it to work within the criteria, digitalize it witha CAD program, and conduct stress and reliability experiments. Because the criteria we have are so specific, we can narrow our focus to create a vessel that will in fact accomplish the necessary goals. Through the criteria, we can test the specific design to categorize each one by their strengths and weaknesses to determine the appropriate design that will accomplish our goals. This categorization will be done through Qualitative Function Deployment. These charts will asses the general needs for each component and thereby allow us to choose the best alternative. Each QFD was prepared through group discussion and prior knowledge of the components.From QFD tabulated results, the design will be constructed and tested within the environment to accomplish the thresholds and objectives defined in the criteria. A tentative set of Quality Function Deployments is presented in the next section.

A basic timeline for our project can be seen in Appendix B.

4.1.1Table 2: Object DetectionQFD

4.1.2Table 3: Movement QFD

4.1.3Table 4: Navigation QFD

4.1.4Table 5: Object Manipulation QFD

4.1.5Table 6: Programming QFD

4.1.6Table 7: CAD QFD

4.2Risk Assessment and Reduction

Table 6: Risk Assessment for Subsystems

The risk assessment above was determined through democratic process among the team members. Each member based the rating of “risk”, or possibility of failure, based on a three point scale for each subsystem based on prior knowledge of the system.

The ultrasonic sensor was an area a high risk. To reduce this risk, the team will work more closely with the sensor and build a prototype. The team rated this sensor as high risk because few had any experience with it.

Wheels were also labeled high because they are inherently less reliable than the other two designs. Reduction of risk can come from decreased use of wheels or the complete disuse with the implantation of an immobile or track structure.

To reduce risk in the conveyor belt object manipulator, a set of plans must first be drawn to show the device in the NXT™ system. A prototype would then have to be built to test its viability in casualty extraction.

1. BUDGET FOR ENGINEERING SERVICES

For this project, three team members and one team captain will work for 18 weeks from initial concepts to final presentation. In terms of hours, each person is supposed to contribute 6 hours weekly. The average pay of a first year engineer is around $43,000 (College Journal 1). For the manager, a minimum of another $10,000 can be added to the first year total (College Journal 1). With benefits, each salary can be doubled and the cost of a work area and facilities can again be taken as double each individual salary. Materials for this simulation project are surprisingly inexpensive at $250 for the Mindstorm™ set. Programming costs are non-existent because we are using a “freeware” CAD program.

Total Cost for the Project:$29,734.00 (Consult Appendix A for detailed calculations)

1. SUMMARY

The problem of disaster mitigation of casualties will be solved in a simulation with the use of the Lego Mindstorm NXT™ system. From the QFD tabulations and risk assessment, we will use a series of sensors to sense the casualty, wheels to move, sensors to navigate, and a latch to manipulate the casualty. We will also use the Lego® programming suite and ML CAD digital drafter.Though there is high risk with the wheels, we feel that this risk can be diminished through proper exploration of its use. The final set-up should successfully complete all thresholds with the possibility of maximum objective completion.

1. REFERENCES

“Better Robots Could Help Save Disaster Victims.” New Scientist. 27 Feb. 2007.

Boyd, Robert. “War and Natural Disasters Driving Push for Smarter Robots.” Digital

Communities. 7 Sept. 2006. 27 Feb. 2007. <http://www.govtech.net/digitalcommunities/story.php?id=100893.

"College Journal: Engineering." College Journal. 5 Jan. 2007. Wall Street Journal. 20

Feb. 2007 <

“Disaster Rescue and Response Workers.”, Dendrite Forest, 9. Mar. 07

1.

Duff, Yim, and Roufas “Evolution of a Polybot: A modular reconfigurable robot.”

Harmonic Drive International Symposium 2001; 2001 Nov 20-21; Nagano, Japan & COE/Super-Mechano-SystemsWorkshop, 2001Nov; Tokyo, Japan.

“Intelligent Sky Division-Report.” ISD Research Areas. 17 Feb. 2006. NIST. 16

Feb,2007http://www.isd.mel.nist.gov/US&R_Robot_Standards/disaster_city/Workshop5_Report-ebook.pdf>.

“Six Limb Disaster Robot.” We Make Money Not Art. 7 June 2005. 27 Feb. 2007.

APPENDIX A: Budget

APPENDIX B: Time Line

Appendix C: Sensor Pictures

(Touch Sensor)

(Ultrasonic Sensor)

(Light Sensor)

- 1 -