Looking Ahead:

If this product were to continue to be developed, what would the next steps be?

  • Increase load capacity of automated shelf to meet that of the manual shelf’s rated load.
  • Continue to improve on safe operation of the shelf.
  • Further develop a multi-shelf design.
  • Increase accessibility of shelf by automating the opening and closing of cabinet doors.

What parts might be redesigned in your product?

  • Eliminating the Arduino prototype board to streamline the control unit into one circuit board.  The ATMega328 8-bit microcontroller itself could still be used and integrated into the single board unit.
  • Increasing the number of accessible I/O connections by using a ATMega2560 microcontroller.  Other microcontroller’s could be used but the operating code would need to be completely rewritten and specs would have to be checked for any differentiation in power requirements and limitations.
  • The current design requires the wiring to the control unit to be directly hardwired to the circuit and Arduino.  An interface to the control unit utilizing a simple plug in from sensor and motor wiring would greatly enhance the usability and ease the installation process.
  • Further research different linear actuators, specifically focusing on their load capacity.  While the current actuator is rated for 100lbs., it struggled to provide the torque needed to lift the shelf from it’s down position and to control the descent speed as weight was added.
  • Improve how the actuator is mounted on the shelf, as it is currently necessary to mount the actuator somewhat loosely on the bolts to allow for rotation.  While the empty shelf would move smoothly through it’s range of motion, jerking was introduced at certain points, especially going through parallel, of it’s descent and ascent as weight was added.  Using a solid ball bearing mount could possibly remedy this.
  • Researching other, possibly mechanical, collision sensing methods is somewhat warranted.

What functionality might be added?

  • A secondary hardwired/wireless push button control, installed at the counter top for example, could be used as a backup to the keyfob in case it is lost or until a replacement for a dead battery is available.  Some users may prefer this as the primary control.
  • Control of additional shelves from a single interface would be the ideal upgrade for this project but also the largest difficulty.  While the current wireless solution can control up to 4 shelves independently, more than 4 would not work even with multiple keyfobs as the receivers are not addressable.  This would result in multiple shelves operating from a push of the A,B,C,orD buttons from any remote.  Initially we had thought to use a smartphone app, scalable to the number of shelves, and communicate with the control unit over Bluetooth.  This may be the way to proceed in the future.
  • Another limiting factor to operating multiple shelves from a single control unit is the added wiring and I/O requirements.  Some sort of wireless unit at each satellite shelf communicating with the base control unit should be explored, but this introduces separate power requirements at each shelf. 
  • An additional actuator can be installed on the other side of longer shelves if needed to keep it from twisting.

What functionality seems unnecessary or is adding unneeded complexity?

  • While some form of collision sensing method is necessary for safety purposes, other avenues should be explored.  While the ultrasonic sensors worked well, the additional wiring, circuitry, and coding might be unnecessarily complex for it’s purpose.  The ultrasonic sensors were chosen for their ease of mounting, and the ability to detect objects without actually coming into contact with them.   A mechanical alternative with high sensitivity to contact would conceivably be acceptable.  One option we thought about was a thin aluminum or plastic plate installed under the shelf with push button type switches at each corner that when the plate contacted an object it would push up against the switch depressing it and stopping the shelf.  Although this would probably increase the mounting complexity, a single I/O could be used to monitor all four switches reducing the wiring and coding requirements.