PDF - Rapid prototyping and testing a visual recognition for Culture Bot - BC-375
Preferred Disciplines: Post-Doctoral fellow in Computer Sciences/Machine Learning or Artificial Intelligence
Project length: Elevate fellowship (2 year program)
Desired start date: As soon as possible
Location: Greater Vancouver Area including Fraser Valley
No. of Positions: 1
Preferences: Preference is for a highly talented individual who could grow into team leader. Language : English
Company: Eleos Robotics, Inc.
Eleos Robotics designs, manufactures, and sells fully automatic weeding solutions for grape growers. Our vision is to one day remove pesticides from thousands of acres of crops by automating crop protection in general with Culture Bot, where “sci-fi and crop protection converge”.
Culture Bot is artificially intelligent. It detects, identifies, and then eliminates weeds that destroy growers' crops. It autonomously patrols crops day and night searching for weeds, and then eliminates them with a high-precision directed energy; it even self-charges with an autonomous charging station and can, therefore, operate day and night without supervision.
We are developing a visual recognition system that will work hand-in-hand with our automomous weeding robot: Culture Bot. The purpose of the computer classification system is to idenify and distinguish weeds. This system will recognize weeds in the evening (with other wavelengths of light) as well as day. Once accomplished, we then eliminate the weed with microwave energy.
This robotic system will be capable of autonomous data acquisition to reinforce its detection and recognition precision. The robot will also be capable of visually navigating and swarming with other autonmously navigating robots.
See website for an overview of our patent pending technology in development.
Background and required skills
- Further develop our existing visual recognition system
- Develop capabilities for robotic systems to be monitored remotely
- Develop cyber-security measures for software systems
- Develop remote/autonomous data acquisiton and reinforced learning systems
- Develop a computer classification system (ANN likely)
- Field test and integrate systems with other Culture Bot systems
- Document development for SRED
- Determine the best strategy and approach to developing the systems (feasibility)
- Set objectives and a work plan, including a GANTT and make hypothesis
- Determine all equiptment needed in the beginning to mitigate lead times and bottlenecks
- Communicate with other members
- Develop a lab version of programming
- Field test and revise programming
Expertise and Skills Needed:
- Machine Learning or Artificial Intelligence
- Neural Networks and Deep Learning
- Data Science
- Various programming languages, Go, JSON, Arduino, C’s, Clib, etc.
- Python is a must
- A bit of hardware experience with robotics is always preferable
For more info or to apply to this applied research position, please