Supports low power IoT cellular and LPWAN networks for rural remote connectivity used in global climate and other monitored area applications

Wilderness Preservation AI Sensor Networks

Our forests, prairies, peatlands, oceans, and glaciers each play an essential role in preserving global biodiversity, sequestering carbon, and keeping our planet in life-sustaining equilibrium. Unfortunately, a high percentage of these resources have already been significantly impacted or destroyed, so protecting and restoring what remains is vital. Machine Learning (ML) at the IoT edge allows readily available sensors to be combined with long-range, low-power wireless connectivity to build smart wilderness monitoring devices with capabilities that can be readily adapted to the various challenges at hand. For instance, identifying the telltale sounds of illegal logging in a forest, monitoring ground moisture, air temperature and CO2 gas levels in areas ranging from peatlands to polar ice caps, or discerning patterns and deviations in migratory birds and animals using ML recognition of unique audio signatures to detect and log movements and presence of species over time and place.


Setting policies aimed at preserving and recapturing vital natural areas is simply the starting point. To have impact, such policies require cost-effective approaches to monitor and enforce to make real progress. The remoteness and lack of human impact that defines these regions also make them some of the most challenging to effectively monitor. Low to zero-impact technologies are needed to fulfill the data insight and policy enforcement objectives without doing more harm than good.

Where SensiML Can Help

By using TinyML sensor models, complex sensor information can be analyzed where it is captured using ultra-low powered solar and long-life battery powered devices that require little or no intervention nor the support of intrusive high-power network infrastructure to pipe volumes of data to the cloud for processing.

Smart Home Energy Conservation

Electricity generation has an environmental impact on our air, water and land. Any reduction in unnecessary electricity consumption reduces our environmental footprint, and any application that can make sense for a single home can have a huge impact if deployed on a large scale. Using AI-based technology, residential HVAC systems, lighting, and appliances can be equipped with sensors that monitor performance, energy use, anomalous behavior, and predictive maintenance cues. Whether an HVAC system monitor, a whole house electrical load analyzer, or occupancy aware smart lighting, the opportunities to conserve power through smart sensor monitoring and guidance can provide notable efficiency gains with simple edge sensor and processing enhancements to new and retrofittable home appliance devices.


Devices and technology aimed at reducing energy consumption usually succeed or fail based on often fragile return-on-investment/payback analysis. Modest gains over time are quickly turned upside down by offsetting initial device costs needed to realize the promised gains.

Where SensiML Can Help

By combining today’s low-cost digital sensors and capable microcontrollers with adaptive machine learning algorithms capable of extracting multi-model insight from low-cost hardware, not only can the energy-saving ROI persuasion become stronger, but also the ability to transform such equipment into smart home goods with multi-benefit consumer selling points. For example, an HVAC virtual mechanic monitoring device can track equipment efficiency, recognize and advise on known fault patterns, and provide new capabilities like multi-zone control leveraging common hardware adders to provide sellable consumer benefits as well as energy savings.

Smart Sensor Aided Disaster Response and Preparedness

Due to climate change, significant weather events are becoming more frequent and even threatening the livability of low-lying regions now subject to ‘100 year’ floods every couple of years. The significant increase in wildfires, floods, and severe weather events ravaging land, properties, and lives demands new solutions to mitigate impact, increase preparedness, and keep first responders safe as they become increasingly called upon to help save lives.


The expanding scope of climate driven disasters demands new solutions that can scale with the dramatically increased scope of coverage needed for both early warning of events and response to those events. Without effective technology to aid the battle, local, state, and federal budgets are being rapidly drained in efforts to keep pace with the rate of change of such disasters.

Where SensiML Can Help

The application of smart edge sensing solutions can play a role on multiple fronts of this disaster response battle. For early warning, remote low-power sensor networks and sensor equipped drones can provide monitoring and critical minutes in identification of wildfires at their outset. First responders can benefit from gesture and activity recognition devices that aid in the coordination and communication in chaotic fast response settings.

Use Cases

Machine Learning (ML) is used to identify the sound of chainsaws and human voices in a forest setting, and then forwards the identification and location information to a smartphone app via a low power, long distance network.

Ready for the Real World?

Developer studies a sensor prototype.
Need to now scale your edge AI project or prototype for an IoT environmental application?

Try Proto Edition which comes enabled by default to generate ML models for motion, vibration, gesture recognition, and countless other subsonic sensor (i.e. requiring <10kHz sample rate) real world applications.