Menu
Our team was tasked to create a design that helps people with Cerebral Palsy, CP, eat. People with CP typically have a hard time using their arms due to uncontrollable shaking. This makes it nearly impossible for them to feed themselves with standard utensils, requiring outside assistance. The main goal of this design is to increase the user’s independence and use minimal outside input.
While people with CP have trouble using their arms to eat, they have enough motor control in their obliques and hip flexors to allow hinging. This could let them lean to a spoon or fork with food. The design would need to be small enough for an easy setup and portable to let them use it outside of the home.
The design must be able to relocate food from storage, a container or bowl, to an accessible location for the user. The design must be portable, so a light weight, battery powered design is ideal.
Concept Design 1 makes use of a 3 DoF arm support. The design would allow a decent range of motion for the user and support their arm as they eat. While this design has no motorized parts, construction may pose a problem as parts will need to be chosen to support varying loads without seizing up.
Concept Design 2 makes use of a 3 Dof sliding platform that is gravity fed. The food would come out of storage onto the slider and be pushed out to the user. The design would present issues with cleaning and accuracy of the sliding arm if the storage container was not properly aligned.
Concept Design 3 vastly simplifies the Obi Robot idea by using a linear actuator to scoop food out of a bowl and deliver it to the user. Instead of a system of multiple servos, this design would use one servo and one linear actuator. It would be limited in the types of food it could deliver.
With these 3 designs in mind, we decided to select Concept Design 3. There were a few reasons that went into this decision. The first being ease of use seeing how the user would only need to push buttons to control the machine. Secondly, it is a pretty simple machine that would deliver the food to a reasonable position in 3-dimensional space for the user to take a bit.
The Extension Bucket collects food from the intrinsic bowl, scoops it against the wall, and extends out to make getting a bite of food easier.
The Extension Bucket has two inputs. The first input is a linear actuator which drives into the food bowl and collects the bite. The second input is a servomotor which rotates the actuator to be parallel with the surface. Once parallel, the actuator extends to full length and stays extended until the user pushes the retract button. The retract button recedes the actuator and rotates it back to the starting position. The Food Bowl in the Extension bucket is mounted to drawer slides on the housing and has a low force spring that returns it to its starting position. The motors are controlled by buttons connected to an Arduino within the housing and their voltage is supplied by a 7Ah battery.
The First engineering analysis pertains to the loading the motor will see. We wanted to determine the torque, power, and current the motor would see when it was subjected to the applied loads to determine the required motor and battery size. The applied loads are the weight of the food, weight of the actuator, and force of the spring.
Assumptions:
The load directions and locations are constant.
The drawer slides apply no friction.
The motor rotates at a constant velocity.
With these assumptions in mind, we derived equations for the torque and power on the motor would need using static equilibrium shown in the figures below.
Arriving at the equations
Once the equations were derived, we wrote a MATLAB script to make calculations and plot the results shown in the figures to the right.
Our calculations show that the motor would see a maximum torque of 6.5 in-lbf or 7.5 kg-cm. The maximum power required to drive the 5v motor at 1 in/s would be 6 W with a current draw of 1200 mA. This tells us that a 10 kg servo should be able to handle the applied loads and a LiFePO4 7Ah Battery would be sufficient to supply it.
The Second Engineering analysis is with respect to the loads seen by the linear actuator. We wanted to determine the axial and perpendicular loads on the actuator to make sure they did not exceed the limits. The manufacturer rates the PA-07 for a static axial load of 6.5 lbf. With the same assumptions made as in the first engineering analysis, we derived equations for the axial and perpendicular loads on the actuator shown in the figure below.
Arriving at the equations
We calculated the results with the same MATLAB script shown to the right. Our calculations show the maximum axial force on the actuator as 0.622 lbf and the maximum perpendicular load as -0.514 lbf. These loads are well within the limits of the manufacturer’s rating. The actuator draws 100 mA with no load. The 0.622 lbf would increase this current to 124.4 mA.
The Third engineering analysis is with respect to the 3D printed parts. We want to ensure that the parts can handle the weight of the servo, weight of the actuator with food, and force of the spring. The actuator weighs 0.3 lb, the spring gives an axial load of 0.7 lb, and that the food weighs 0.125 lb. We applied these forces to our 3D model in fusion and used the static simulation to determine minimum factors of safety for our 3D printed parts. The simulation results for factor of safety and Von Mises stress are shown to the right.
Actuator Sleeve
There are many loading situations on the actuator. Specifically, there is a different load case for each angle. The current angles are 40°, 5°, and 0°. Using these angles, the maximum stress occurred at the 5° angle where the spring force and weights of the food and actuator are all present. The minimum factor of safety is 15 and occurs where the servo will be mounted. We then loaded the actuator 0° with 5lbf in each direction to simulate a user taking a bite of food. The maximum stress occurred when the force was applied in a downward direction. The minimum factor of safety is 15 and occurs where the servo will be mounted.
Spoon
The simulation for the spoon is to make sure it can handle the spring force and weight of the food. The minimum factor of safety is 10.36 and occurs where the spoon will be mounted onto the actuator.
Spring
The simulation for the spring catch is to make sure it can handle the max spring (3 lbf) and the spring force during use (0.7 lbf). The max spring force occurs when the drawer slides are fully extended. This is much further than the actuator will push the spring. The minimum factor of safety of the spring catch during use is 8.236 and occurs at the base of the spring catch. The minimum factor of safety at max spring force is 2.76 and occurs at the same location. Since the spring catch will not see the max spring force often, a factor of safety of 2.76 is acceptable.
Before we began building the housing and food holder, we wanted to make sure that we could control the servo and actuator.
The first two images show the servo being controlled with a normally open button, arduino, and power supply.
Servo in Neutral Position
Servo Bottom Position
We then wrote a script to control the actuator using a relay, power supply, and an Arduino. Shown in the following images.
Actuator Circuit
Extended Actuator
Retracted Actuator
We then wrote a script to control the actuator and servo at the same time, and power them both with a battery.
Actuator and Servo rest position
Actuator and Servo end position
After we knew our code could work, we built the food holder, housing, and support arm. Some changes were made to the housing to make it easier to assemble. We added in rectangular slots to the housing panels so they would fit together. We also changed the support into a rectangular tower that would hold the servo and actuator.
Food Holder.
Housing being glued together.
Housing and Food holder assembled.
Circuit to be built.
Tower and Circuit being constructed.
Circuit mounted inside housing.
Full Assembly
After the assembly was completed, we noticed that the spring we had chosen too strong for our application once friction was involved.
Spring connecting food holder and housing.
We replaced the spring with a much weaker rubber band, and it seemed to fix the issue.
Rubber Band connecting food holder and housing.
We then made sure our Voltage regulator was outputting 5V as intended.
Voltage Meter testing the output from Voltage regulator.
5V output
We then began the first test run of our assembly. Upon plugging in the battery, the servo went to the wrong position. It went to roughly 45° above parallel instead of 40° below parallel. When we pressed the button, it traveled to roughly 85° above parallel instead of parallel.
Servo 45° above parallel
Servo 85° above parallel.
We altered the code, and the servo began working as intended. However, the servo could not make it through the full range of motion due to the spoon getting stuck on the bottom and back of the food holder. The servo would start stuttering while trying to reach the 40° position requiring us to elevate the actuator.
We chose a strong servo because we knew that this could be an issue, but it seems like either the servo is not getting enough current or that it is getting caught. We believe that it is getting caught. Future revisions could use different materials to greatly reduce the friction but for now, we reduced the friction by removing rough spots and increasing the speed the servo traveled through the holder. This seemed to mostly fix our issue.
The first successful run showed some flaws in our spoon’s design. The side walls of the spoon are too shallow to contain some foods and the spoon itself is not deep enough. Future designs should account for this.
Note: If actuator and servo are not returned to their initial positions during start up the servo will snap to its initial position which could cause harm to the user.
Our project focused on prototyping a portable low-cost feeding device for people with cerebral palsy. We constructed the housing, tower, and food holder out of plywood. 3D printed the spoon, food bowl, and button mount out of PLA, and chose electrical components to suit our needs. Our chosen concept design required many revisions to make a useful prototype. The end product is portable and mostly functional with a few key revisions needed to make it fully functional. The prototype retrieves food and delivers it to the user with the push of a button.
