Jyosh Integrated Agriculture Robot
Introducing automation in Cotton-Cultivation in India for the first time
Project
Background
Working as the only Industrial Designer in the team, I designed the first visual concept model of Jyosh Integrated Agriculture Robot for a Mumbai-based startup 'JYoSh Ai Solutions'. From October 2020 to April 2021, I established a successful concept design, iterated the product structure, and refined several versions of the design. Mainly targeting Cotton Cultivators of India, I considered the Indian farming standards. As planned, an actual size working prototype is already made and tested currently.
​
Achievements:
-Winner at Agri India Hackathon 2021
-NASSCOM recognition
-Shark Tank India round finalist
ROLE
Industrial Designer
METHODS & TOOL
-Survey, Secondary Research, Competitor study
-Solidworks, Blender, Illustrator
DATE
October 2020 - April 2021
PRACTICES
User Research, Material Handling, Prototyping, Product Design
TEAMMATES
Sharad Chandra Lohokare (Founder CEO)
Jitendra Ahirrao (Founder CTO)
Sachin Nikam (Software Head)
Sambhaji Jadhav (R&D Head)
My Role
-
Built mathematical models in Solidworks & executed functional tests of 3d
printed and machined prototypes to meet the functional requirements of the robot. -
​Created the official 3D animation video of the robot vehicle on Blender software
-
Designed the pitching video of over 3 mins for the company for their Shark Tank
India Finale. -
Developed the company's corporate logo
1.jpg)
Context
India is in the top 3 cotton producer countries of the world. But, the productivity and profitability is far less than many nations. Losses incurred due to weeds, insects and diseases account for a substantial proportion but the major drawback is the labor-centric cotton cultivation in India. Mechanization is limited to the use of tractors and motorbikes. Harvesting ready cotton fields do not get enough labor to work at the right time. Farmer suicidal rate and depression are serious concerns caused due to lack of profitability.
​
​
In other countries
Faster/ Efficient/ Profitable
In India
Slow/ Inefficient/ Over-working
Indian agriculture system is still largely dependant on a manual workforce at every stage of farming. This places many kinds of limitations to develop ahead and compete for efficiency.
Back in Mumbai, Jyosh Ai Solutions Start-up dived deep into this problem and hired me as an industrial designer to work along with their team. To start with, we tried to understand why the advanced vehicles and robots of other nations cant operate on Indian fields.
Learn - Define - Design - Build - Revise - Deliver
Process
We followed a lean startup methodology to focus on the pivotal functions in the first prototype. The design process started by conducting user research and studying extensively the competitor models. We defined the problem and started breaking it into tasks and subtasks to begin the ideation phase. Right now, the prototype is being tested and further iterations will be made after feedback.
Research
Problems associated with Mechanical picking of cotton in India
-
Sympodial BT hybrid cotton balls in India are not suitable for mechanical picking
-
Small size holdings of Indian farmers
-
Indian cotton plant height grows up to 6 feet which is not ideal for the current machines
-
Pre-cleaning of trash content will be a major operation
-
Initial large cost of imported machine
-
Harvest aids are not used to a lot extent in India
Current issues with manual cotton picking
-
The availability of laborers is uncertain
-
Unsynchronised cotton-picking due to weather conditions
-
Lack of time to complete the operations
-
Low income and poor health conditions of farmers
-
Less cotton gets picked up
These were the research findings after studying design patents, past articles and papers on cotton picking mechanization, Cotton boll varieties, automated machine categories and their mechanisms.
How might we design an integrated agriculture robot which is small and cost-friendly and is ideal for the cotton variety of India?
Design Goals
With the goals divided into smaller tasks and activities for simplicity, a chart of dependant and independent tasks was formed.
System study
A rough system architecture was devised after organizing the cotton harvesting activities in proper steps.
In addition to the cotton-picking system, weed removal and pesticide dispensing system were also a part of the robot. The functioning systems were connected with a charged battery to power them. We collected Cotton images by visiting the fields and studying actual samples. Forming a thorough Cotton database was imperative for the software to work.
Ideation Phase
Starting with quick line diagrams and continuous brainstorming sessions, I took initiative to construct a 3D model on computer software Solidworks. It suddenly made many relations and variables clear. Important measurements included the width between cotton plant rows, the effective area of one plant, and the distance between the plants in a single row.
We visualized the green columns as the Cotton plants and adjusted the space between the rows accordingly.
Design affordances and tolerance added more depth to the visual. After 20-25 revisions, a good representation was formed.
It was a rigorous phase in forming an entire model with a rough placement of different components.
Design Challenges
1. As many key components of this system were not standard parts, it was challenging
to refine the entire model virtually without designing the small components first.
​
2. Placement and animation of various sub-assemblies on the same model proved
to be difficult due to technical limitations of the software.
To solve the challenges, we started designing the key components of the Cotton picker mechanism, especially the robotic arm which picks the cotton, in detail. Followed by animating the entire system on Blender software more effectively.
Prototyping
It was necessary to test the cotton picker mechanism in real-time for its validity. I spent many days at workshops and 3D printing units to get samples made of the designs and test them.
Animation
By the end of 7 months, I delivered a running animation video of the entire automated vehicle of the first-ever practical concept. The current team is conducting usability tests of the prototype and will iterate based on the feedback.
(The video and Render images are subject to Copyright and property of 'Jyosh Ai Solutions')
An actual prototype under testing
Animation Video
Reflections
1. Small team = Heavy workload
It was my first experience as the only designer in a start-up team. Although it was a tonne of workload, I tried to go step-by-step and document each change. I maintained a timesheet of regular activities, tasks pending and completed. My role required me to prototype the design and test it, and not limited to just designing a 2D visual.
2. Internal Communication was the key
I spent quite an amount of time discussing even minor changes to understand the concepts thoroughly and be on a common consensus with the team. I learned to effectively convey my rationale and justify the decision in regular meetings. Lastly, proper communication helped to boost the project and maintain the right direction.