Traptic pitch @TCD2019

Очень интересный питч компании Traptic, которая производит робототехнику для уборки клубники и других ягод и фруктов (farming robots). На TechCrunch Disrupt 2019 в Startup Battlefield компанию представлял Lewis Anderson, Co-Founder & CEO. Переводить не будем, чтобы не потерять аутентичность. Обратите внимание на мастер план и на цифры.

Manual labor. 200 years ago, we picked wheat by hand. Every single stock. A bunch of people out in the field with hand tools, doing things manually. When we invented combines. Machines harvest quickly, efficiently. And that enormous volume. But they like the delicacy and the precision necessary to harvest most crops. So people still pick more than $200 billion worth of fruits and vegetables by hand each year. So why is this a problem? Take strawberries. 20% of the strawberries grown in the US rot in the fields because there’s nobody available to pick them. Labor is also super expensive. It costs $15,000 per acre, or about 25 cents per pound. Just for manual labor for harvest. The flavor is getting even more expensive and more scarce due to an aging workforce around the world, national political dynamics and California Labor Law. Introducing Traptic – giant farming robots pick strawberries. Our robots are saving the world’s food production system by automating the work people don’t want to do. 

Our robotic platform drives across the field while our vision system detects the berries and figures out how to reach them. Then our patented gripper delicately extracts the produce from the plant. So what does this actually look like? Switch to the camera feeder the machineThe systems camera take a snapshot of the scene. Our vision system detects the berries in that image, decides which to pick, figures out the position of that berry and determines the structure of the plant and bed around them. Then it plans a path to grab it and our software commands the robot arm to move the gripper to the berry. Our gripper then lifts the berry off the field and closes around it firmly but gently. We then command the robot arm to yank the grip or away from the plant. And because we have a firm grasp on the berry, the stem pops off.We finished by depositing the berry on the conveyor.

Видео выступления Льюиса и ответы на вопросы

Just in case you thought this is a lab experiment, let’s play that video. Our system does all this in real time, wall hanging off the back of a tractor and moving continuously over real strawberry fields. It is robust to dust, dirt, rain, wind, vibrations and the other aspects of the tough farm environment. And It operates all day, every day for months and years on end. Our gripper is totally unique and patented. This gripper must extract berries from a crowded environment, hold them at a certain angle to help to bend the stem, apply a great deal of force to break the stem of the berry. And do all this whole being very delicate. We grabbed the berries with soft food safe silicon belts that we mold around different materials. This gives a gripper mechanical compliance, allowing it to grab objects of varying size and shape without needing any information about the objects properties. We close the fingers with a defined amount of force, allowing us to grab objects of different sizes without crushing or dropping them. This lets the same gripper grab produce across a large range of size and shape without needing any information from software, or any hardware changes. Built on a robust data set Trapter robot can find berries and determine ripeness with equivalent accuracy as people. And finally, it does all this locally. no internet, no offsite compute. Industry standard machines are not designed for delicate produce. Conversely, some startups target low value activities like weeding, only worth one to $2,000 per acre. Or they harvest fruits like apples, where they’re only needed a few months a year. Traptic focuses on year round strawberry harvest. We don’t require any change in growing process, providing a smooth path for deployment. 
Our master plan. Get 3 billion in revenue by picking strawberries. 20 billion in revenue by picking the other fruits and vegetables, then go after $200 billion in market in crop value by automating and improving remaining activities. Robots as a service, We pick strawberries, and growers will pay us 23 cents per pound.

Вот что такое хороший мастер-план

This is slightly less than the current spend. We provide all aspects of the service: machine, tractor driver, etc. Traptic has developed testing partnerships over the last three years with the largest commercial growers including nature ripe and Driscoll’s and has done daily testing over the past six months. Traptic will deploy commercially next year. We built big fixed wing autonomous airplanes for an AVI si University project. Then Ben and Brian went to UTC aerospace systems. And I went to Microsoft. After studying traffic we recruited from UC Berkeley, University of Michigan, SpaceX, GM, and Driscoll’s. Our advisors are globally recognized industry pioneers search bill on Jay Nathan Darren, Peter bill. The future of robotics is outdoors, Join Traptic and help save the world food production system.

Bonus: интервью 2017 года с Льюисом о будущем робототехники:

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s