Final project proposal: Mauricio

Haptic feedback and actuation for power tool control

Objective

The overarching purpose of my project is to improve haptic feedback mechanisms in order to provide an intuitive and precise way of controlling large scale power tools with hand gestures. The inspiration comes from the dominion of sculpting, taking into consideration the wonderful precision sculptors reach by modeling materials with their own hands or with hand-tools. This kind of precision is not possible in a “direct way” (meaning modeling the piece in real time such as in clay modeling) when dealing with large scale pieces or difficult materials to work with. On the other hand, mass production tools such as industrial robotic arms with several degrees of freedom might be able to achieve this tasks, yet it is my opinion that pieces produced in this way lack a “soul” component to them. The purpose is then to allow the artist to have a very “expressive” way to control a robotic arm, ir order to leverage both the power/precision of the machine but also the uniqueness and “happy accidents” of a human produced piece. The thesis is that this “very expressive way” may be achieved by gestural control of the machine, mimicking as much as possible the original contact between hand and materials. Examples of gestural control of robotic arms are readily available, yet they always (in my opinion) seem to address extremely simple tasks such as grappling objects, moving them from point A to point B, and similar. The reason I think most systems are not capable of achieving greater precision is because gestures are performed in free air, neglecting the tactile feedback that direct contact gives us, and for which millions of years of evolution have prepared our sense of touch and muscles. My intent is to test a method of adding haptic feedback to a gestural control mechanism for a robotic arm and performing user studies to determine if it enhances or not the interaction.

Scope

It has been discussed with the faculty leading this course that the project will be broken down in pieces, because of its complexity and also to better adjust the academic requirements of different courses. Hence in “Making Things Interact” the mission will be to create a simple feedback & control mechanism not for a full fledged industrial robot, but for a much simpler tool, for example a drill press or similar. The system must allow for:

• Control of some aspect of the machine
• Haptic feedback

Implementation plan

I will:

• Either embed a glove with an Innertial Motion Unit (IMU) for position or acceleration/rotation tracking and also with vibrators for feedback, or get something equivalent in the market, something that I can hack.
• Create a few degrees of freedom control with the glove for a simple power tool, e.g. control the height of a drill press by actuating over its lever. The actuation might be as simple as with a motor and a piece to attach the axis to the lever of the drill.
• Create a control loop so that the glove sends data to the actuator and in turn measures data that allows it to show basic haptic feedback. For example, the current of the motor actuating on the lever may be measured, hence determining more/less torque, which is related to the head of the drill press touching or not the material being drilled. Certain current level will trigger the vibrating response of the glove, completing the feedback loop.

Notes

A way to embed a high end IMU and vibrating feedback into the glove in a convenient form factor, timely and tidy manner may be by attaching a smartphone to it, hence this avenue will be explored.

Since the example input is acceleration/rotation of the hand and the intended feedback vibration, the feedback itself might make the input to the system really noisy. This may be:

1. Exactly what needs to happen (think of when you are going really hard against a material with a power tool: it may vibrate in a way that makes it difficult to retain precision)
2. Not desirable (based on user assessment). For this project I think this could be addressed by having 2 gloves, one for vibration and the other for input, or filtering the input in a way that the vibration is omitted.

Deliverables

• Glove with vibrating feedback and sensors (does not need to be IMU based but that is the plan for now) for controlling an actuator for a power tool.
• Working code, schematics, documentation, etc.
• Demo