Crop scouting,
from the ground up.
Mark Robotics is building Bibi — a small crop-scouting rover for high-value agriculture. Bibi captures row-level video in structured crop environments and turns it into simple AI-assisted scouting reports.
Crop scouting is still difficult to scale.
Growers often rely on repeated manual inspection to spot plant stress, disease signals, fruit load, ripeness variation, and other problem zones. But scouting every row frequently can be time-consuming, inconsistent, and hard to document.
Manual inspection takes time
Walking every row at the right cadence is hard. Teams stretch thin during peak season, and coverage is often the first thing to give.
Problems can be missed between rounds
Early stress, isolated disease pressure, or uneven ripeness can slip through between scouting passes — especially across long blocks.
Existing tools may lack row-level detail
Drones and satellites give a useful overview, but rarely answer the questions that need a closer, repeatable look at the canopy.
Bibi turns field video into scouting insight.
Bibi is designed to collect repeatable field footage and highlight areas that may need human attention. The goal is not to replace growers or agronomists, but to give them a clearer view of what is happening across rows.
Capture structured row-level video
Bibi is designed to move along the row at a steady pace, recording the canopy from a consistent height and angle.
Detect plant and fruit signals
Footage is processed to flag visual indicators: stress, damage, fruit load, ripeness shifts, and gaps.
Generate a simple scouting report
Outputs are summarised by row and block — the kind of view a scout might write up after a walk.
What we’re exploring
We are validating which of these use cases creates the strongest value for growers first.
Spotting visible stress patterns and damaged foliage as they appear, row by row.
Plant stress and leaf damage
Estimating the count and distribution of clusters or fruit set across each row.
Fruit load and cluster density
Surfacing zones that are progressing faster or slower than the surrounding block.
Ripeness variation
Aggregating small signals into block-level views of where to send a human first.
Problem zones across rows
Built for high-value crop environments.
We are currently speaking with growers across several crop types to understand where Bibi should be focused first. The common thread is structured environments where frequent visual inspection affects quality, yield, and timing decisions.
Where we are now
We are early. Below is an honest snapshot of what is in motion and what is still ahead. Bibi is an early prototype — not a commercially available product yet.
Customer discovery
Speaking with growers and agronomists to map how scouting actually happens today.
Hardware prototype
Early rover platform under development. Focused on row navigation and stable capture.
AI video analysis demo
Building a first pass at detecting plant and fruit signals on captured row footage.
Feedback & pilot sites
Looking for interview partners and future pilot sites to validate the approach in the field.
We’re speaking with growers and farm managers.
We are not selling anything yet. We are looking to understand how crop scouting works today, where manual inspection is painful, and what kind of tool would actually be useful.