Research

Title: Energy Consumption Rate in a DC Motor

How mechanical work affects the performance of and energy consumption rate in a motor—attached to a robot for example—is an ongoing engineering question. It is well understood that negative mechanical work done on a motor would increase the current through the motor and decrease the voltage across it. As a result, the energy consumption in the motor would increase. To understand this phenomenon, consider a circuit where a motor is connected to a battery of voltage V and draws a certain current I. The power consumed would be the product V×I. The motor draws less current when there is no mechanical load than when a negative mechanical work is done on the motor. This is because the motor runs faster without the load and therefore a significant amount of back electromotive force (EMF) builds up, and this back EMF sends a current opposing the current supplied by the battery, thus reducing the total current going through the motor. However, when a mechanical load is added to the motor or when the motor is forced to slow down by an external force, then the back EMF is smaller and therefore the circuit draws more current. Although the effect described above could easily be demonstrated with a DC motor, an ammeter, and a battery connected in series, a methodical, quantitative study of the power consumed by a motor as a function of time is in order. Our objective in the present work was to investigate in detail energy consumed by a DC motor. Using two different current sensors, Adafruit’s INA219 and Sparkfun’s ACS712, we have measured the current through and voltage across a DC motor, run without a mechanical load, with a load for a brief period of time, and completely stalled for a brief period of time. With both sensors, we have measured the load voltage using the Arduino Uno microcontroller. We have documented the relevant circuit diagrams, the Arduino code in its entirety, and the correct experimental procedures.

Figure: Experimental set up for current measurement

Research

Title: The Effect of Hemp Concentration in Natural Fiber Composites

This research addressed the use of natural fibers in the creation of composite materials. Composite materials are formed when two or more materials are combined to create a new material. The goal is to create a new material that maximizes the strengths of the individual materials while minimizing their weaknesses. Using natural fibers allows for the creation of a new material that contains easily accessible, natural materials and is cost efficient to create. Our research focused on combining hemp fibers with epoxy and a hardener in 1-, 5-, and 10-percent concentrations. Once combined, tension testing was done to determine and compare their average modulus of elasticity and ultimate stress. This allowed us to determine which concentration of hemp composite was the most rigid and which could sustain the most force before breaking. It was determined via graphs of stress vs. strain that the modulus of the 1% samples was on average 616.8 MPa and had an ultimate stress of 36.33 MPa, which was the largest value for the concentrations we tested. The 5% samples showed a lower modulus of 573.33 MPa and an ultimate stress of 20.18 MPa. The 10% samples had the largest modulus of elasticity of 928.47 MPa and an ultimate stress of 21.35 MPa, which was much less than the 1% sample, but only slighter higher than the 5% sample. The stiffer a substance, the larger its modulus of elasticity. The 10% hemp concentration samples were the most rigid because of the strength of the hemp fiber. We expected the 10% hemp concentration samples to also have sustained the most force before breaking, and thus the highest ultimate stress; however, the 1% showed a higher ultimate stress. This could be based on the fact that as the hemp percentages increased so did the amount of air bubbles, which caused inconsistency in the uniformity of the samples and therefore increased the weakness of the 5% and 10% samples.

Figure: Tensile strength of 1% hemp composite samples

Research

Title: The Feasibility of Growing Medicago sativa in a Transparent Hydrophilic Medium

The versatile alfalfa (M. sativa) was used in this study to examine the sustainability of plant life in a hydrophilic transparent medium. To create the transparent medium we used aquabeads, a hydrophilic copolymer of isobutylene and maleic anhydride. Aquabeads, capable of absorbing ×200 water, has a similar refractive index as water so it can become transparent. Thus aquabeads is ideal for creating a transparent medium that makes it possible to observe the growth of plant roots. M. sativa was used in this study because of its versatility, low water consumption, and its high nutritive and ecological benefits. There has been relatively little research on the sustainability of transparent media for plant growth. Alfalfa sprouts were grown in containers of the transparent medium, while monitoring light intensity, pH, and temperature. Alfalfa grows best under low light intensity. A successful alfalfa growth in a transparent medium would have tremendous application in green design. Green roofs on commercial buildings may help to create high ratings on LEED (leadership in energy and environmental design) certifications, which may lead to a higher property value. Alfalfa has high aesthetic value due to his fast growth and high biodiversity- and may contribute tremendously to the LEED ratings of commercial properties.

Figure: Testing light intensity during alfalfa growth in transparent medium

Research

Title: Tracking and Distance Detection Methods for a Humanoid Robot

Humanoid robots are robots with resemblance to qualities/features of the human body. These robots have two arms, two legs, a torso, etc. Some humanoid robots even have facial features resembling that of a human. Cellularly Accessible Expressive Semi-Autonomous Robot (CAESAR) is a humanoid robot in development at NYU-Polytechnic School of Engineering’s Mechatronics and Controls lab. Currently, CAESAR’s eyes consist of two cameras that are being used for tracking and distance detection of various objects. Open Source Computer Vision (OpenCV) is being used to design a program that allows CAESAR to find and track an object of a certain color and shape in its field of view, based on the user’s request. The computer processes images that it collects from the camera; it then communicates that information to an Arduino. The Arduino Mega 2560 then communicates with the Arbotix-M to drive the motors to its desired location.

Research

Title:Teaching Emotion Recognition to Autistic Children with CAESAR

Autism Spectrum Disorder (ASD) is a developmental disorder typified by repetitive actions, impaired communication and social skills, and delayed language development. Often autistic children fixate on one particular object in their environment and exhibit a heightened fondness for mechanical toys that move in predictable, repeated patterns. Children with ASD experience difficulty with self-expression and are overwhelmed by the subtleties of human facial expressions vis-à-vis the relatively simpler interaction with mechanical objects. Currently robotics is being explored to help engage autistic children to learn language skills and to develop empathy through caring for a robotic “pet.” However, robot therapy options to engage autistic children to learn emotion recognition and emotion expression skills remain to be fully explored. Our work focused on Cellularly Accessible Expressive Semi-Autonomous Robot (CAESAR), a humanoid robot in development at the Mechatronics Lab of NYU Polytechnic School of Engineering. In particular, we worked on enabling the robot to make recognizable facial expressions that can be recognized by autistic children so that the skill developed with the robot can transfer to social interactions with other humans. An iPad app was designed to enable the children to select one of six specific emotions. Touching one of the expression icons enables CAESAR to enact the corresponding emotion dynamically, i.e., using facial expressions and arm gestures that reinforce the emotional state being portrayed. The emotions were based on psychologist Paul Ekman’s research into non-verbal communication and his “Six Universal Emotions”, the most immediately recognizable facial expressions that he observed among international cultures. The expressions of happiness, sadness, anger, surprise, fear and disgust are Ekman’s universal emotions that are incorporated in our app. CAESAR and the emotion recognition app enables autistic children to strengthen emotion recognition skills by repeated exposure to the same facial expressions. The designed platform can also be used to help autistic children identify and express their own emotional state given their language deficits. CAESAR and the app are unique among the few emotion recognition options available among autism robot therapies in that the children can select the emotions that the robot demonstrates, allowing them to communicate through CAESAR in addition to building their social skills.

Figure: Emotion prototypes

Figure: CAESAR is happy(left), angry (middle), and afraid (right)

Research

Title: The Enhanced Development and Testing of the Abaxial Appendage of a Bio-mimetic Platform for Fish Mating Behavior

We formulated and tested three hypotheses. Hypothesis I: The displacement of a fish replica due to simultaneous linear motion along the x and y-axis, results in the fish replica moving in a spatial probability defined as an ellipse. Hypothesis II: The vertical, z displacement of the support rod due to the linear actuator, while the rod undergoes simultaneous linear motion along a horizontal x and y-axis, results in a spatial probability defined as a parabolic cylinder. Hypothesis III: A biomimetic replica can be used to stimulate live fish and the analysis of resulting fish behavior can show differences resulting from physiological or genetic variations. The experiments to test the first two hypotheses were conducted. The components were designed in Solid Works (SW). The whole platforms, including the new appendage, were assembled in SW. Computer simulations and mathematical analyses were performed in SW. Videos of the virtual functioning of the platform and the motors constituting the structure of the platform were analyzed using the Pro-Analyst software. Manual analyses of the geometry of the models for the fish trajectories were also performed. The plastic components of the arm were fabricated by means of the Dimensional Elite industrial 3D printer. A Makerbot printer was also used. The major part of the platform was constructed with 2 cm × 2 cm aluminum t-slotted frame. A program code to operate the platform was written using the Arduino software. The platform was assembled and the motion of the appendage with the fish replica recorded as videos. By means of the ProAnalyst software, the videos will be analyzed. We were able to demonstrate that the experimental data were consistent with our hypotheses.

Figure 1. The Bio-mimetic platform with its appendicular appendage