We finish what the paper started.
The methods behind our systems are published, cited, and open to inspection.
Why this page exists
Anyone can claim a research lineage. Very few companies will show you the work.
Everything below is either published, patent pending, or measured on data we will describe. Where a number appears, we can tell you how it was produced and on what.
Books by our team
Carlotta A. Berry. Mobile Robotics for Multidisciplinary Study. Morgan and Claypool, 2012.
Carlotta A. Berry and Brandeis H. Marshall, editors. Mitigating Bias in Machine Learning. McGraw-Hill Education, 2024.
Patents
More than fifteen patents pending across AI and agricultural technology, including autonomous drone agrobotics.
Methods and evaluation
How we report results, so you can judge them:
- Performance is stated against a held out set defined before training, not against the data the model learned on.
- Field results are reported with the conditions attached: lighting, weather, motion, sensor, and platform.
- Failure modes are published alongside the successes. A system with no documented failure modes has not been tested hard enough.
- Where a figure comes from a third party, we cite it. Where it is ours, we describe the setup.
Collaborate with us
We work with university labs, agency research programs, and industry partners on funded research, SBIR and STTR teaming, and field trials.
Talk to us about research collaboration