Robots in Space! – Alicia Kavelaars, Founder and CTO, OffWorld – How Advanced Machine Learning is Expanding Capabilities for Robotic Workforces in Extreme Environments

Alicia Kavelaars, Founder, and CTO of OffWorld (offworld.ai) discuss her company's efforts to build robotic technology for extreme environments, with advanced machine learning and AI, to assist humans to be offworld. Offworld, roughly defined as 'away from earth' is the core point of Kavelaars' company's mission—to enable "smart robots working under human supervision on and offworld, turning the inner solar system into a better, gentler, greener place for life and civilization."

While experimentation in space is costly, Kavelaars explains that much of their testing is done in more financially practical extreme environments such as underground mines and seabeds, etc. OffWorld's initial prototypes focus on mining in extreme environments and will be tested in underground mines first, but these robotic workforces will be extremely helpful in a space environment for colonization preparation efforts. Kavelaars' robot workforce is already designed for an end-to-end mining solution. From mining through processing, OffWorld's robotic workforce is capable of handling the job.

OffWorld's robots can go deeper underground, to mine in ventilation and extreme temperature conditions that humans could not withstand, and of course robots can work 24 hours a day for continuous workflow. With thousands of robots in a workforce, Kavelaars expects to move humans to supervisory roles, out of harm's way, and allow advanced machine learning to continue to enhance the robots' effectiveness. OffWorld's robotic workforce is fully autonomous, end to end, but if a robot in the workforce experiences a contingency situation in which it cannot make the assessment on its own, human supervisors step in and handle it.

Kavelaars explains how OffWorld's robots operate independently, in that they are not centrally programmed with a hive mentality, but perform as solo operators with a heightened awareness of the other robots working alongside them. This design helps to alleviate potential single-point failures as Kavelaars' robots will recognize failures and independently, but with a collective understanding, support any one-off failures.

The robotics founder expounds upon her desire to take her robotic workforce to space and the preparation to make them fully autonomous in advance of their mission. Kavelaars' team employs machine learning to help robots evaluate precise techniques for proper mining, to allow the robots to vary their approach based upon arising conditions, as opposed to simply inputting algorithms that would not allow for adjustments.

Additionally, the CTO and founder discuss her company's plans to introduce robotic workforces for other areas of industry as well, such as large construction projects. As robotic workforces can work independently and autonomously, but can also work in tandem with humans, robots can significantly increase efficiency in many industries with challenging environments. And as machine learning advances and sophistication of robotic technology is refined even further, robotic workforces will be able to master practically any industry-type tasks