A couple of weeks ago, I had the pleasure to accompany a team from CyberTracker and Harvard University visiting the Rio Bravo Conservation and Management Area in Belize to field test PAWS, a Predictive Patrol Planning application integrated with SMART.

Part of the Transboundary Maya Forest in Belize, the Rio Bravo Conservation and Management Area is a 260,000 acre reserve. This is the largest private protected area in Belize, protecting 4% of the country’s land area and home to 392 bird species, 70 species of mammals and 22 vegetation types.  The reserve has been using SMART to assist their patrols since 2018 with support from WCS Belize.

Hill Bank Field Station in the Rio Bravo Conservation Area, Belize (Credit: Monica Harris/SMART)


“Before SMART, we used GPS, pen and paper to record all patrols and infractions. It was hard to capture everything during and at the end of a long day patrolling.” said Baltazar Garcia, Field Station manager at Hill Bank, one of the two stations in the reserve.  “We can now collect all the information with SMART due to customization of the data model to respond to our needs. This allows me to produce timely reports to all collaborating agencies and donors”, explained Ramon Pacheco, Manager, Admin and Planning for Programme for Belize, the conservation organization who owns and manages the reserve.

Head Ranger, Marcos Corado, recording an illegal logging incident in SMART Mobile (Credit: Monica Harris/SMART)


Protection Assistant for Wildlife Security (PAWS)

The years of patrols logged in SMART made this reserve a great field testing site for PAWS. The Protection Assistant for Wildlife Security (PAWS), developed in the lab of Milind Tambe, Professor of Computer Science in the John A. Paulson School of Engineering and Applied Sciences at Harvard University, uses machine learning to predict hotspots for potential illegal activities. PAWS measures patrol effort – how often and how much time rangers spend in the area – using a 1km grid, and gathers information on other features, such as: topography, hydrology, animal density, patrol posts, roads, rivers, vegetation cover, precipitation, etc. All of these data are combined with information on illegal activities recorded in SMART in order to make risk predictions.


Lily Xu and Sonja Johnson-Yu, Harvard University academics who are leading the PAWS work, described to the reserve team that PAWS has been tested in other sites in Africa and southeast Asia with very successful outcomes. They demonstrated how the predictions can be refined to focus on specific illegal activities, such as illegal logging, or even a specific season, providing additional information to Protected Area managers to assist their decision making.


“PAWS is designed to provide rangers and protected area managers with additional insight to augment their expertise,” said Lily Xu. “A key objective of our field visit to Rio Bravo was to understand their priorities and constraints on the ground, particularly as this is our first site in Latin America, so the wildlife, terrain, and traditions all differ from what we’ve seen before.” 

Harvard team discussing the PAWS riskmap for Rio Bravo Reserve with the rangers and reserve managers (Credit: Monica Harris/SMART)


We visited some of the high risk areas shown by the predictions, some of which were already well known to the rangers and some off trail. During our patrol the rangers recorded illegal activities, such as: baits placed for hunting and evidence of illegal logging. Lily Xu and Sonja Johnson-Yu reflected that this field visit has given them greater insight into the management of the reserve, the nature of the illegal activities and additional information that can be added to the model to improve the prediction.


This field visit was a huge learning experience for us. Not only did we get to understand the operations surrounding ranger patrols, but we also had a chance to have a dialogue about PAWS’s strengths and how we can continue to grow and improve it. Previous field tests have shown that PAWS works well for predicting the presence of snares'', said Sonja Johnson-Yu. Now, in a landscape where there is no snaring and it’s primarily illegal logging and poaching with firearms, we have a whole new set of challenges for PAWS that we’re looking forward to tackling.


A beta version of PAWS is available with SMART 7 for testing and the SMART and Harvard teams will continue collaborating to refine the tool and predictions.


Left: Bait placed by hunters to attract paca (Agouti paca) . Right: Ranger Jose Manuel showing an area of illegal logging. (Credit: Monica Harris/SMART)


SMART Mobile

This visit was also a great opportunity to field test some of the newer features in SMART Mobile and hear feedback from the Rangers in the reserve. The latest build of SMART Mobile to be launched with SMART 7 has a number of improvements, including navigation features such as go-to, ability to manually add or drop a GPS point, a statistics page that shows the patrol progress, ability to add photos and audio to observations and many other features. Justin Steventon, co-founder of CyberTracker, the software behind SMART Mobile, was part of the team visiting the reserve and joining the rangers on patrols. 


“While SMART Mobile builds on the foundation of CyberTracker, it is an entirely new system designed with ideas from the broader SMART community. It was exciting to see just how proficient and natural the rangers have become with it in the relatively short time they have been using it. While chatting with them, I was impressed by how their feedback spanned from ideas for simple improvements all the way to entirely new kinds of features. Overall an enlightening trip which will help to make SMART Mobile an even more valuable tool for conservation.”

Justin Steventon, CyberTracker co-founder and SMART Mobile developer.


Left: Ranger Elias Romero using SMART Mobile (Credit: Monica Harris/SMART). Right: SMART Mobile new navigation screen including go-to (Credit: Lily Xu/Harvard)


Overall this was an amazing experience for everyone. Thank you so much to Programme for Belize, Hill Bank Rangers and staff and WCS Belize for the hospitality and for making this trip possible.


See you in our next blog, signing out!


Monica Harris

SMART Program Manager