Close
0%
0%

Hempbot-Bee

Hempbot-bee is an agricultural robot to help make the hemp farms boom!!!

Similar projects worth following
Hempbot-Bee is a mobile robotic system with a combination of deep learning neural networks, feature detection and vision based object avoidance to seek out and perform task to reduce the cost and maximize the profit of hemp farming for CBD oil.

With legalization, hemp shows promise of being the new boom industry on the edge. Analysts are predicting a $200 BILLION industry by 2024. Hempbot is an agricultural
robot to help make the hemp farms boom!!!

In the Commonwealth of Virginia, where I live, hemp became legal to grow with a license in March 2019. In only a few months, a gold rush has started. With CBD oil producing plants, the possibility of tens of thousands of dollars per acre is now possible.  With the possibility of big profits come big challenges. Hemp for CBD oil is extremely labor intensive, but with these challenges comes business opportunities to solve those challenges. 

Any good product needs to fill a need. From talking to new local hemp farmers the following list of needs are the first to come to mind.

  • Planting/transplanting from a greenhouse
  • Cultivating weed control
  • Insect control
  • Watering
  • Applying fertilizer/feeding the plants after they leave to green house
  • In greenhouse care
  • Data logging every plant. Detailed records of every plant for state records.
  • Trimming/collecting of buds

With many different needs (opportunities) Hempbot-Bee will be a modular design with a variety of attachments. Not only are there business opportunities in supplying a bot but also for supplying different attachments for the bot.

With a lot of different task to be performed on a farm, different sizes of bots will be required. I plan to build several different bots to perform the different task. The bots will be networked to allow tracking the different jobs and collect a database of each plant. Hempbot-Bee will be the first of several bots. Bees are a very inspirational creature. Housekeeping, collecting nectar and pollen, taking care of the brood and guarding the hive, Hempbot-Bee will be a worker bee in the network. Hempbot-Bees responsibilities will be weeding/cultivation, watering, feeding the plants and collection visual data of the plants for tracking and the requires government records.  

  • The Eyes have arrived!

    Dennis08/04/2019 at 13:43 0 comments

    I finally got the two cameras for the stereo vision. Now I can start testing the feature matching and ranging with two cameras instead of just jpegs images.

  • Feature matching for object avoidance (Another brick in the wall :) )

    Dennis07/28/2019 at 20:08 0 comments

    In order to have Hempbot avoid running into objects with a vision system, I’m planning on taking a screen shot from one camera, detecting objects then finding the same objects in a screen shot from the second camera. This will allow the distance from the object to be determined. Here’s the result of trying to find a picture of a brick in a picture.   

    Here’s the picture of the brick

    The results are in red.

    So far, not bad.

  • First Parts arrived!

    Dennis07/24/2019 at 21:29 0 comments

    Cool the first parts are here. Time to start the build….love starting a new project!

  • Raspberry PI and mobilenet ssd test using Python

    Dennis07/18/2019 at 00:13 0 comments

    Testing Deep learning on a Raspberry PI 3 B with a mobilenet ssd model. So far so good. The PI can recognize my wife’s flowers as plants. The DNN is working surprising well with just a cheap USB camera I bought off of Ebay. This seems to work regardless of the light level. The one problem is the PI is very slow, about 2 seconds on each test frame. I may try an Intel compute stick.

    This was one indoor test for fun. The cheap camera also worked well in low light.

View all 4 project logs

Enjoy this project?

Share

Discussions

Similar Projects

Does this project spark your interest?

Become a member to follow this project and never miss any updates