Research Notes‎ > ‎

Teapluckingsimulation

CONFIDENTIAL  between PDM and IITMANDI

PENDING AGREEMENT ON IP ISSUES
16 May 2013

authored by paola di maio
shared with TAG,

Tea Plucking Robots

This is a very schematic outline for a brief 'joint research proposal', put together by PDM following a
visit to Palampur last weekend  (10 May 2013) with TAG and Venket, and conversations with the Director of the Bioresearch 
Park who suggested he may be interested in collaborations. This draft is currently shared to a restricted list, to obtain early feedback and a minimal agreement on IP, before being formally proposed to IITMandi
faculty for open collaboration, if the proposal is of interest.


The problem
============
Automating, or enhancing the manual process of tea leaves plucking is an open challenge
Currently it is possible to use machinery to pick tea leaves, but this results in lower grade crop, so the tradeoff between using manual vs automated plucking results in a quality/cost considerations that impact  the market [1]

A possible ethical issue attached to the impact of automation on the  labour market force is noted [2] However at this stage, the proposal is exploratory and focuses on tackling the purely technological, socio technical and scientific aspects of the problem/challenge, by integrating different disciplines and technologies. Ethical consideration of industrial applications of any outcomes of the research will be
taken into account, if relevant, at deployment stage (if we get there)

Goal and expected outcomes
======================
To explore state of the art, general intelligence approach to tea plucking automation, Expected novel outcomes are
- a model of expertise
-  possible prototypes
- experimental results (publications)
- if successful, a patent

Methodology
============
An interdisciplinary research framework is proposed.
A state of the art review of the field of contemporary approaches [7]  carried out 
In the first instance a physical 'expert model' is attempted, by studying observing and modelling the skill and expertise of tea pickers, including body gestures and behaviours. The EM aims to to capture the complexity and multidimensionality of the subtle interactions between the human and the plant, a number of modelling techniques to be explored (sensorial, visual, cognitive etc). It is envisaged the outcome of different modelling approaches may be integrated to create the richest possible model.
Secondly, thanks to integration of new manual skills capture technologies, such as 3D Sensing, [3] the  essential behaviours modelled in the EM are captured.
Thirdly, combining the lastest sensing techniques and sensitive robotics, devise the prototypes.

A number of hypotheses can be formulated and tested using a combination of the three steps above.


Scope
=====
1. Literature review and state of the art, feasibility and viability
2. Iteratively researching and prototyping
3. possibly patenting? researching the application scenarios


Process
=======
This could be a faculty wide project, open to any member of faculty to collaborate, where different 

approaches may be tested. The documentation and contribution would be maintained in an openly editable 

document, however, any IP that may accidentally be generated during the collaboration, would pertain to the  authors/contributors as per the  respective claims/statements in the project document.  Or other arrangement to be discussed/agreed among project participants. Publications may arise from the project will be authored by all contributors


Resources required
=========

Some 3D sensing technologies, cooperatio with tea plucking fields, more, TBD



[1]http://flora.coa.gov.tw/view_eng.php?id=180
[2] http://www.bbc.co.uk/news/magazine-20542255
[3] http://www.gesturetek.com/3ddepth/introduction.php
[4]http://www.kurzweilai.net/tactile-sensor-for-better-human-prostheses-personal-assistive-robots
[5]http://www.gatech.edu/newsroom/release.html?nid=210251
[6]GESTURE SENSING
http://www.newscientist.com/blogs/onepercent/2013/01/hands-on-with-leap-motion.html
[7] http://www.msrsas.org/docs/sastech_journals/archives/sept2011/14.pdf
Comments