A thermal drone is great gear for a variety of missions — from surveillance and security patrolling to industrial asset inspection. And there’s been plenty of new model releases over the last year. 

But with a sharp price tag, the “cheapest” models start at $6K, so you don’t really want to wing it (pun intended). Learn how to choose the right thermal drone from our quick guide. 

Important Thermal Drone Features To Consider 

To choose an all-around pleaser, check each option against these criteria: 

Thermal Camera Settings 

Many Electro-optical (EO)/Infrared (IR) camera payloads for drones are available. Look for systems with a resolution of at least 640 x 512 px for crisp images. 

A broad spectral band in the IR sensor is advantageous for better scene recognition and performance under various weather conditions (e.g., fog, rain, snow). Check if you can set custom isotherm ranges manually to further tune your equipment for the use case.

The best thermal imaging drones combine thermal sensors with an RGB camera for an impeccable shooting experience.

Radiometric Functionality

Thermal drones with radiometric features capture precise temperature readings, rather than differences in ranges. This allows you to calculate precise data in measured area (min, max, average °F/ °C), giving a broader read of trends. For example, you can evaluate heat distribution across industrial pipelines to measure thermal efficiency.

Some thermal drones can be auto-programmed to focus on specific temperature ranges for streamlined data collection. For example, you can set a custom range to inspect temp variations in cold storage rooms to detect heat loss.

Gimbal Stabilization

A solid gimbal reduces blur in thermal footage, especially models with gyroscopic stabilization. It also provides extra angular velocity to track fast-moving targets 

(e.g., if you’re using a thermal drone for border security tasks). 

Flight Time and Range

Most enterprise quadcopters can stay airborne for 30 to 50 minutes, depending on weather and load. Larger fixed-wing and VTOLs like Albatross UAV can cruise for up to 4 hours at 20 m/s (and it can be equipped with a thermal camera payload). 

Range matters for covering larger areas. Advanced drone transmission systems can sustain a steady range of up to 6-9 miles (10-15 km).

For extended operating time, you can choose a tethered thermal drone, which can hover for days when connected to a power source. With Bavovna’s AirTower Mode, tethered drones can operate fully autonomously even in GPS-denied environments.

Environmental Durability

The best thermal drones boast exceptionally high endurance, including wind resistance up to 23 knots, IP55 rating against water and dust damage, and built-in redundancies. For safe missions, look for models with redundant properrels, INS components, and motors.

Best Thermal Drones for 2025 

Need some recs? Here are the top picks from Bavovna’s team:

• Skydio X10. Measuring just 31.1” x 25.6” x 5.7”, Skydio X10 can stay in the air for up to 40 minutes with a max speed of 45mph. The hybrid imaging system combines a narrow 64MP RGB camera, a 48MP telephoto one, and a radiometric thermal camera with  640 x 512 px resolution and under 30mk sensitivity. 

• Autel EVO Max 4T. With an IP43 rating and an extra pair of hot-swappable batteries, Evo Max 4T is a reliable companion for a range of missions. Equipped with a hybrid RGB/thermal camera, this UAV can muster an impressive temperature Measurement Range of -20°C to 550°C. GPS-denied navigation mode is a great bonus.

• Inspired Flight IF1200. IF1200 model from Inspired Flight is sturdy and robust. It can lift up to 19.1 lbs in payloads while staying in the air for 35 to 43 minutes. It’s compatible with the Gremsy VIO F1 thermal camera, featuring a 4K zoom sensor, a 640×512 radiometric sensor, and an integrated 2400m laser rangefinder. Thermal sensitivity range is ≤ 50 mk, giving you crisp imagery under any flight conditions. 

Discover more UAV companies in our directory. 

Collecting aerial data used to be a daunting challenge. Tape measures, foot patrols with a theodolite, or cost-inhibitive helicopter flights. 

Unmanned aerial vehicle (UAV) mapping changed the aerial data collection game, bringing extra speed, lower costs, and greater precision. 

New to the concept? Here are the essentials you need to know about UAV mapping. 

UAV Mapping and Surveying Use Cases 

Fixed-wing UAVs, VTOL drones, and enterprise quadcopters can stay in the air for up to an hour (and sometimes more), giving surveying teams ample time to perform various geodesic tasks. The best mapping drones also include specialized payloads for high-precision data collection like thermal cameras, multispectral sensors, magnetometers, gas detectors, and LiDAR systems for 3D scanning. 

Here’s how businesses use UAVs for surveying and mapping tasks:

• Topographic mapping. Drones help create HD orthomosaics and 3D models for cadastral surveying, allotment planning, and a host of other civil surveying use cases. Swiss Canton of Valais used a WingtraOne mapping drone to conduct mountain village land surveys in 3 days, instead of 2 weeks with conventional methods. 

• Mining exploration. Using drones, operators can assess resources and plan excavations based on geospatial information. Rugged, in-door models also help assess sub-terrain corridors to ensure safe and effective operations. WACO S.R.L. used the Elios 3 drone to inspect dangerous rock detachments inside its quarry (Italy), providing teams with valuable operational data. 

• Urban planning. City planners rely on UAVs to collect visual data for 3D modeling, land classification, and smarter resource allocation. Thanks to automated route planning and high-precision data capture, drones substantially reduce the cost and fieldwork hours. A surveying team in Weinan City, China, used drones to collect oblique cityscape imagery with greater efficiency. Based on this data, a comprehensive 3D model was created with an accuracy level of up to 5 cm.

• Road construction surveying. For large-scale transportation projects, UAVs provide seamless data capture for large-area linear maps, reducing the complexities of planning, monitoring, and documenting new construction projects. The Norwegian Public Roads Administration uses drones to survey underway projects more effectively. Mapping a 3-mile road with a drone can take just 2.5 hours and $270 in labor costs vs 6 days and $5,200 with terrestrial laser scanners. 

Shortcomings of UAV Mapping 

Although UAV mapping comes with a slew of benefits, it’s still a tedious process, susceptible to different disruptions. 

Weather can be a major factor as lighter, commercial models are inoperable in high winds, heavy rain, or snowfall. Fog, in turn, can cause sensor interference, increasing the risks of collisions and data capture accuracy. 

Terrain type can also aggravate the UAV’s technical limitations, leading to signal loss, mapping errors, and scrambled navigation. Dense vegetation, large water bodies, mountains, sudden elevation changes, and high-rise buildings, limit GNSS/GPS signal propagation. In such cases, it’s worth looking into a solution for GPS-denied navigation. 

Regulations. Many UAV aerial mapping use cases require BVLOS permissions, which may be hard to obtain in certain jurisdictions due to bureaucratic red tape. Drone operators must also comply with privacy requirements, as well as other rules related to flying over people and close to restricted areas. 

Data accuracy. Although drone technology has made major leaps, technical limitations still remain. Automated, on-device data processing can impact footage quality and accuracy. Also, discrepancies between vertical measurements can vary significantly without Ground Control Points (GCPs). On average, you need to place 12 GCPs for small to medium sites 

(7 and 39 ha) and up to 18 for the large sites (342 ha). This adds extra workload for field teams. 

Improving UAV Mapping with Bavovna Hybrid INS 

Overcome the challenges of GPS signal obstruction with Bavovna Hybrid INS Navigation Kit. Measuring just 150 x 134 x 73 mm, Bavovna kit enables reliable, long-range navigation with AI-powered sensor fusion. Custom-trained for each drone and a variety of operational scenarios, Bavovna helps operators fly mapping missions without getting held back by signal propagation delays, interferences, or drift bias. 

Discover Bavovna Hybrid INS Navigation Kit. 

An internal reference unit (IRU) is a three-axis system that provides precise attitude, velocity, and navigation information to the vehicles using internal sensors. Unlike basic sensor arrays, an IRU applies sensor fusion technologies to correct errors and estimate PNT with higher precisions, enabling seamless navigation in GPS-denied environments.

Components of Inertial Reference Unit (IRU)

IRU processes data from internal sensors to provide navigation, stabilization, and autonomous functionality to piloted vehicles. Depending on the sensor combination, an IRU can continuously measure angular acceleration, linear velocity, attitude (roll, pitch), position,  platform azimuth, magnetic and true heading, body angular rates, and more. 

The most common IRU sensor components include: 

• Gyroscopes: Measure rotation rate around three axes —  roll, pitch, and yaw

• Accelerometers: Mesure linear acceleration to determine changes in speed and position 

• Magnetometer: Measures the orientation relative to the Earth’s magnetic north to determine heading direction  

All collected raw data is processed on the edge to optimize output accuracy with Kalman Filtering — a probabilistic data fusion algorithm that corrects for noise and drift to provide more accurate PNT data to the vehicle. IRU can also perform dead reckoning calculations to enable navigation in GPS-denied environments. Processed data is immediately transmitted to the flight controller through SPI, I2C, or CAN for real-time course correction. 

IRUs are the key component of many autopilot systems in commercial aviation and maritime vessels. 

Inertial Reference Unit (IRU) vs Internal Measurement Unit (IMU): What’s the Difference?

An internal reference unit and an internal measurement unit are the main components of modern internal navigation systems (INS). But each serves a slightly different purpose. 

• An IMU only aggregates raw internal sensor data without applying any extra computations. 

• An IRU collects and processes raw data to create precise position, velocity, and orientation outputs. 

In most cases, the IMU plays a supporting role. It provides extra inputs for redundancy or slight course correction when GPS is used. IMU data can also be combined with visual inputs from cameras or LiDAR to enhance simultaneous localization and mapping (SLAM) algorithmic outputs. Or, in the case of Bavovna, used as part of an AI-powered navigation solution. 

IRU, in contrast, may work in conjunction or independently from other external systems (e.g., GNSS) to provide a higher degree of autonomy and reliability. But that also comes at a higher price tag. Honeywell’s internal reference systems, used in many private and commercial jets, have a starter price of $300,000. Each also weighs anywhere between 9 and 40 pounds.  This inhibits IRU’s usage for UAVs. 

Bringing the Power of IRU to Drone Navigation 

The main advantage of IRU is on-edge data processing which enables more advanced, autonomous aerial navigation scenarios. But that comes at a ‘cost’ of larger hardware size because commercial IRUs include large processors, memory storage, and sometimes redundant system components. 

IMUs, in turn, are optimized for lightweight applications like drones, which come at a ‘cost’ of limited processing powers. At Bavovna, we’ve decided to solve this problem with the help of AI. Rather than trying to enable continuous on-device processing, we’re pre-training AI algorithms locally on the vehicle flight data. Effectively, we’re helping UAVs develop ‘memory’ and then use it autonomously in the field. 

For example, we’ve successfully trained a model for AirTower flights — fully autonomous ascending and descending to set height, hoovering, and returning to a designated landing point without any GNSS connectivity with a 0.5% positioning error. Similarly, we trained other drone models to fly more complex trajectories with 98% accuracy. 

Discover how Bavovna enables high-precision aerial navigation with an AI-powered INS solution. 

Uncrewed aerial vehicles (UAVs) have become the crux of military operations and security patrols. But every drone reconnaissance or patrol mission comes with the almost imminent danger of equipment loss as UAV jamming has become the norm in contested territories. 

How UAV Jamming Works 

All UAV models have one inherent weakness — communication links. So attackers use various techniques to disrupt communication between a drone and its operator or navigation systems by causing signal interference and/or overwhelming signal receivers. The usual targets for jamming are 900 MHz, 2.4 GHZ, and 5.8 GHz frequencies, as well as 4G and 5G frequencies. 

Common UAV jamming techniques include: 

• Radiofrequency (RF) jamming pollutes the same frequency drone operators use with powerful noise signals or rapidly switching frequencies to meddle with more advanced drone models. 

• GPS jamming or spoofing: Jammers overwhelm the UAV’s drone receiver, causing it to lose its position data. GPS spoofing is a more advanced technique, used to mislead the onboard systems drone about its location and force it to head elsewhere. 

• Broadband jamming aims to overpower the UAV’s communication systems by flooding the airspace with high electromagnetic noise to disrupt the UAV’s control link. 

• Narrowband jamming targets specific coms frequencies, used by drone operators to neutralize the communications. Unlike broadband jamming, there are fewer ‘collateral impacts’ on other airspace users. 

GPS jamming, in particular, is on the rise, especially in the areas of ongoing military conflicts. But the problems also ‘spill’ to other regions. Since the Russian invasion of Ukraine, GNSS jamming and spoofing have increased substantially across the eastern Mediterranean, Baltic Sea, and Arctic regions, according to the European Union Aviation Safety Agency. 

Drone jamming is also on the rise all across the US-Mexican border, where drug traffickers rely on jammers to thwart drones, deployed by the border security forces. Given the ease and low cost of obtaining UAV jamming tools (a device can cost a couple hundred dollars), the question of protection becomes equally important for civilian and military use. 

Levels of GPS interference, recorded on March 23, 2024. Source: GPS World 

How to Counter UAV Jamming 

Various anti-jamming options have emerged to protect the ‘weak link’ in UAV devices—

However, the best protection is removing the underlying vulnerability. GNSS/GPS technology isn’t just susceptible to targeted jamming. Signal also degrades due to natural magnetic interferences—complex geological terrain, high-rise urban structures, and the natural levels of emissions, produced by various equipment.  This complicates drone use in mining, telecom, oil and gas, and many other industries. 

Internal navigation systems (INS) have emerged as an alternative UAV navigational technology. Advances in AI sensor data fusion make INS as reliable as GPS-only navigation. 

Bavovna has developed a low-energy, external navigation system, compatible with most commercial drone models. The devices process data from onboard sensors (accelerometer, gyroscope, compass, etc) and optional external systems (e.g., computer vision camera or LiDAR) with the help of pre-trained AI algorithms to provide reliable navigation in GPS-denied environments. Positioned in an EMI-protected case (which successfully passed EMF tests), Bavovna hybrid INS offers staunch protection against UAV jamming. For extra security, our system can be integrated with SIGINT RF modules. 

Boasting a longer range compared to other internal navigation systems for drones, our system maintains a 98% average accuracy rate, even when flying complex trajectories. All thanks to fine-tuned AI models, trained on live vehicle data, which can compensate for individual sensor deficiencies. 

Discover how Bavovna is securing drone operations with a hybrid INS system. 

Since 1993, the GPS has been tightly integrated into our daily lives. From recording bank transactions to guiding transatlantic flights, the technology generates about $1 billion a day in economic impact. 

But just like any other system, GPS has its fair share of limitations. Challenging terrain, signal jamming, and spoofing can render it useless. And that happens quite a lot. Drone use cases in mining are severely limited by naturally occurring magnetic interference. Thousands of commercial flights get affected every year by targeted or incidental GPS signal jamming. Not to mention countless security and military operations, where signal spoofing is the name of the game. 

Source: FT

Soundly, alternative technologies exist for GPS-denied navigation — and here are the top 3 solutions. 

Hybrid INS Powered by AI 

Internal navigation systems (INS) rely on data gyroscopes and accelerometers to estimate the vehicle’s current position, in relation to its last known point. The problem, however, is that many off-the-shelf systems lack accuracy, especially over a longer range. 

Bavovna is changing that with its AI-driven inertial navigation unit. Compact, low-power, and EMI-shielded, Bavovna brings AI sensor fusion technology to UAV navigation. The onboard unit can process data from any number of sensors—accelerometer, gyroscope, compass, barometer,  vector airflow, ultrasonic, infrared, or optical flow sensors—to deliver high-precision navigation in GPS-denied environments. The endpoint positioning error is just under 0.5% even when flying complex routes. Our solution is fine-tuned on live flight data from your vehicle, ensuring unbeatable reliability and durability in a variety of conditions.

For instance, our latest deployment on Aurelia X6 Max Pro-D allows performing fully autonomous air tower missions—vertically take-off, hoover, and land, without any maps or additional correction from GPS or RTK.  With Bavovna’s AI kit, you can safely establish communication relays, perform terrain reconnaissance, perform security monitoring, and fly a range of other missions without worrying about GPS signal degradation, jamming, or loss.  

Quantum Positioning Systems 

The Royal Navy is looking to another emerging technology to improve INS—quantum computing.  Atoms exhibit quantum behavior changes in response to the smallest amount of motion when cooled near absolute zero. These changes can be measured and used to obtain positioning, navigation, and timing (PNT) data. The catch? Cooling down atoms requires huge, power-hungry equipment. 

Aquark Technologies may have found the answer to this quandary. The quantum startup develops miniature quantum systems. Its compact cold atom navigation system uses lasers to bring the temperature down to (-273.15C), which makes it possible to collect motion data on an atomic level and use it for navigation. Aquark Technologies has been successfully tested on a Royal Navy patrol vessel in October 2024. 

Silicon Photonic Optical Gyroscopes (SiPhOG)

Fiber-optic and ring laser gyros offer the best accuracy, but they are also too expensive for many commercial applications. MEMS gyroscopes are way cheaper but lack precision. ANELLO Photonics wants to close this gap with its SiPhOG technology. 

A silicon photonic integrated circuit is used to manufacture the waveguides on-chip, allowing the company to achieve Fiber Optic Gyro performance with a standard silicon manufacturing process. Its INS system has a drift rate of less than 0.5° per hour and demonstrated strong performance in GPS-denied environments. It maintains accuracy within 0.1 m over distances of 0.8 km without GPS, even in orchard environments with limited GPS signals.

Navigation technology moves quickly and better alternatives to GPS are emerging every day. Many also boast high customization like Bavovna’s AI navigation kit, allowing multiple deployment scenarios across different hardware — fixed wing, tilt wing, VTOLs, multi-copters, and FPV drones. Contact us for a free demo!

Internal navigation system (INS) uses motion and rotation sensors and an onboard computer to determine the vehicle’s position, orientation, and movement speed without using visual references.  

Originally developed in the MIT Instrumentation Laboratory for a B-29 bomber in the 1950s, INS has become a staple for self-contained navigation for aerospace, maritime, and automotive industries. 

How Does an Inertial Navigation System Work?

INS uses dead reckoning to determine the vehicle’s current position by using its last known coordinates as the starting point for comparison. It then provides real-time position and navigation data by correlating changes in starting point, speed, and direction with new sensor inputs.

Most modern internal navigation systems include an inertial measuring unit (IMU) — a sensor subsystem that provides raw data inputs like altitude, position, orientation, angular rate, and linear velocity.

Source: ResearchGate 

Inertial measuring units (IMUs) typically feature the following sensors:

• Accelerometers to calculate changes in velocity and position

• Gyroscopes for angular velocity estimation to detect rotational motion

• Magnetometers to determine movement direction relative to the Earth’s magnetic field

• Barometers/Altimeters to measure atmospheric pressure for altitude calculations.

Common Types of Internal Navigation Systems 

INSs differ significantly in hardware configuration—each having different tradeoffs in accuracy, cost, and application feasibility. The common INS types are:

• Strapdown Inertial Navigation Systems (SINS) have sensors strapped directly to the vehicle. They’re lightweight and easy to implement, ideal for drones and light robotics. But SINSs require high computation power to handle sensor noise due to vehicle motion.

• Gimbaled Inertial Navigation Systems (GINS) use gimbals to ensure greater reference stability. But they’re heavier, more complex, and susceptible to mechanical wear.

• Fiber Optic Gyro-based Inertial Navigation Systems (FOG INS) leverage fiber optic gyroscopes for precise rotation measurement. FOG is more immune to vibration and environmental interference but costlier. 

• MEMS-based Inertial Navigation Systems feature accelerometers and gyroscopes, based on Micro Electro Mechanical Systems. They are cost-effective and compact but have lower accuracy than FOG or RLG systems.

• Ring Laser Gyro-based Inertial Navigation Systems (RLG INS) use ring laser gyroscopes for precise motion measurement. They boast high durability and are vibration-immune, but come at a premium price.

• GNSS-Aided Inertial Navigation System typically features a 3-axis gyroscope, a 3-axis accelerometer, and a GNSS receiver (and sometimes a 3-axis magnetometer) for navigation. Each contributes different coordinates for high accuracy. The problem? If the GPS is down or lagging, navigation becomes unreliable—and that’s a major limit industries aim to solve.

Inertial Navigation System vs GPS: What’s the Difference? 

INS is a self-contained system that doesn’t require external connectivity (e.g., satellite or wireless networks) to guide the vehicle. As such, it’s less prone to magnetic interferences or targeted attacks, especially with EMI shielding.

GPS, in turn, is a satellite-based system that provides positioning data only when and where there’s an unobstructed line of sight to satellites. This makes it unsuitable for underwater navigation, UAV or aircraft flights in contested environments, or autonomous driving through tunnels or underground shafts.

Due to GPS’s vulnerability to signal loss, interference, and jamming in contested environments, many organizations use INS over GPS. Recent advances in sensor fusion and AI improved internal navigation system accuracy and connectivity range.

Getting more from your INS with AI

Bavovna developed an Al-enhanced INS for uncrewed vehicles that delivers 98% accuracy over a long range and supports fully autonomous flights in GPS-denied environments.

We’ve developed a compact strap-down model with an IMU and AI-powered flight control, weighing only 800g. For navigation, we apply sensor fusion to accelerometer, gyroscope, compass, barometric pressure, and airflow data, with the option to connect more sensors. Each AI model is custom-trained for your uncrewed vehicle on at least 100 hours of live flight data to ensure top accuracy during autonomous flights.

In the field, our hybrid INS system can maintain under 0.85-meter deviation of single-point positioning without GPS, RTK, and optical navigation at 500 meters altitude with 18 m/sec wind. Learn more about Bavovna AI Navigation Kit.  

GNSS signals are inherently weak and susceptible to interference. Highrise buildings, industrial machinery, and high-voltage powerlines, not to mention intentional signal jamming, frequently meddle with UAV navigation. 

At Bavovna, we’re building customizable AI navigation kits, combining rugged, low-energy onboard hardware and custom navigation models for fully autonomous flight in GPS-denied and EW-threatened areas.

Our latest release is AirTower mode, ensuring fully autonomous vertical climb, stable hover, and controlled landing in signal-restricted areas.

Introducing AirTower Mode 

AI AirTower flight mode enables a UAV to autonomously ascend to the set height, hover, and return to a designated landing point without GNSS or maps. The endpoint positioning error during field tests was about 0.5%. The system also supports autonomous take-off, landing, and return to home with a single point positioning ±1m latitude.

In the default implementation, the model uses data from the accelerometer, gyroscope, compass sensors, barometric pressure, and multi-vector airflow. We can also integrate data from ultrasonic, infrared, optical flow sensors, or computer vision systems.

The latter is optional. Compared to computer vision, H-INS is less dependent on environmental features, not susceptible to motion blur, and consumes less power. Field tests showed that sensor data fusion suffices for most navigation scenarios.

AirTower mode was originally developed and tested on Aurelia X6 Max Pro-D, one of the best enterprise drones in its category. It can also be adapted to a wide range of other UAV and ROV models. 

Each AI algorithm is fine-tuned using live data from the drone, making the system universally applicable for VTOL, multi-rotor, or fixed-wing drones. At 800 grams (1.7 pounds) and low SWAP (peak consumption of <12A), our system is compatible with lightweight quadcopters with limited payload capacity. AI AirTower mode is useful for tethered drone operators seeking fully autonomous operations in any condition.

AI AirTower Mode Use Cases 

It’s easier to visualize the benefits of any technology in a real-world context. Here are four operating scenarios in which Bavonva’s AirTower mode adds the most value.

ISR missions

Maintaining a controlled vertical climb and stable hover is essential for various intelligence, surveillance, and reconnaissance (ISR) missions.

Our first use case was border security. EMI-protected UAVs can provide a vantage point for detecting illegal crossings or suspicious activity, replacing laborious foot or vehicle patrols. Port authorities and coastal guards can also benefit from our navigation system, which maintains high accuracy even in areas without distinctive landmarks like water reservoirs or sand dunes.  

EMI-shielding allows Bavovna-enabled UAVs to safely operate in hospital environments for terrain resonances or FOB defense scenarios. For example, a UAV can periodically ascend to monitor threats and auto-land in a hidden, designated spot to minimize detection. Our system is fully payload agnostic. It can be easily integrated with a radio repeater and various transmitters. For example, SIGINT RF module for proactive reconnaissance of EM threats and autonomous bypass of EW/EM obstacles for extra safety. Or even a portable, counter-UAS EW system. 

Radio Relaying

Drones offer benefits like rapid network deployment, extensive coverage, and real-time data collection. Combined with a tethered or large-battery drone, Bavovna’s systems can aid in establishing communication relays in remote, rugged, hospital, or disaster-affected locations.

Full autonomy in GPS-denied environments means you can deploy drones in mountainous, forested, or dense urban areas without signal interference. The autonomous drone can help (re-)establish communication links between dispersed personnel and the command center for better coordination for rescue, supply, and tactical operations. After the mission, it can land at a designated spot for easy retrieval.

Emergency Response 

The main challenge of disaster management operations is achieving the fastest response, especially for rescue efforts where every second counts. Bavovna’s AirTower autonomy mode allows deploying drones almost immediately in affected areas to get a good aerial view and provide ongoing situational awareness to first responders.

One scenario we had in mind is wildfire detection. Dense forestry is challenging to operate in and remoteness means poor signaling. Bavovna-equipped rescue drones can transmit exact site coordinates to ground teams for investigation. In Turkey alone, drones have helped successfully detect over 2,000 wildfires last year. 

Site and Asset Inspection 

Specialized inspection drones are also becoming a staple in asset management programs. Aramco uses VTOL drones to survey miles of pipelines, spread over large desert areas. South Korean SK Telecom deploys quadcopters to detect loose bolts and nuts on antennas. Mining companies use drones to automate terrain data collection for better stockpile management.

In all of these cases, reliable connectivity is crucial, but not always readily available. High-voltage powerlines generate strong electromagnetic fields, that can interfere with the GPS receiver. High EMI, generated by telecom towers, also leads to signal delays or loss of, especially if the drone is flying close to the asset. Bavovna’s solution helps effectively overcome all of these challenges. 

Fly Autonomous Missions Without Signal Loss 

When GPS is reliable, all goes well. When it’s not — the flight outcome is uncertain. With Bavovna’s H-INS system, you can take mission security out of luck’s hands and into your control. 
Easily activated from the Bavovna Mission Planner, AI AirTower mode navigates you through GPS-denied and EW-threatened areas reliably. Compared to alternatives, our system provides up to 98% higher positioning, navigation, and timing (PNT) accuracy at a range of up to 30 km / 18.6 miles. Get in touch to learn more about all features.

An inertial navigation system (INS) is the key enabler for stable UAV navigation when the GNSS signal is degraded or unavailable — and that happens a lot. High-rise buildings, dense vegetation, electromagnetic emissions from nearby structures, or targeted UAV jamming frequently cause signal loss and costly equipment loss. 

That shouldn’t be your case, however. Recent advances in sensor fusion and artificial intelligence (AI) enable drones to autonomously navigate complex droughts in GPS-compromises and denied environments. 

Here are the top four AI navigation solutions for UAVs, available on the market right now. 

1. Bavovna 

Bavovna offers a versatile AI navigation kit, compatible with multi-rotor, VTOL, and fixed-wing models. The low-SWAP system includes IMU and Al-powered flight control, encased in an EMI-shielded, carbon case (which has been field-tested for reliability). For extra protection, SIGINT RF modules can be added to the INS to identify and overcome EW/EM threats. 

The Airtower use case needs to be covered.

Bavovna H-INS got a proven 0.85-meter deviation of single-point positioning without any GPS, RTK, and optical navigation on board, at altitudes 500 meters with a wind up to 18 m/sec.

Unlike other INS, Bavovna’s solution can provide complete autonomous navigation in GNSS-denied environments using a proprietary sensor-fusion AI algorithm. The algorithm is already pre-trained on standard sensor data (multivector aerial speed, multi-vector air flow, aileron feedback, barometer, compass, gyroscope, and magnetometer) to ensure high positional accuracy and situational awareness. Then further Bavovna fine-tune the model, using approximately 100 hours of flight data from a particular airframe to ensure the best possible performance. 

Bavovna H-INS stays sensor agnostic and AI-driven sensor fusion can be easily extended to integrate extra navigation aid systems, lidars, or computer vision. For example, if you also want to use simultaneous localization and mapping (SLAM) for mapping the terrain or to use object detection for your applications.

The hybrid approach currently allows Bavovna to provide ultra-low end point positioning error (EPPE) of under 0.5% at a range of 30 km / 18.6 miles — a substantially longer range than non-AI INS and radio-based navigation systems can deliver. 

Competitive Strengths 

• No GPS, RTK, RC ground, or network signal is required to pilot UAVs using the Bavovna AI navigation system to perform autonomous take-off, Return-To-Home, and landing.

• Sensor fusion algorithm fine-tuned for each customer’s flight gear and application scenarios.  

• Modular, scalable unit design with low power consumption, protected with composite radio shielding.

2. George Autopilot 

Source: uAvionix

George autopilot system by uAvionix is a robust UAV navigation system, based on certified DAL-C hardware and CubePilot autopilot architecture. Both proved their mettle in a range of enterprise UAV operations. 

To power safe flights, George 2 replies on data from the IMU units and military-grade geomagnetic sensors. You also have the option to add detect and avoid (DAA) protection by adding pingRX Pro or ping200X Mode S Transponder. On the downside, George doesn’t use SLAM or odometry to provide extra situational awareness, unlike Bavona.ai.  

Another advantage is that the entire navigation system weighs only 80 grams (2.8 oz). It’s also optimized to minimize power consumption for longer flight times and hardened to withstand abnormal power conditions, lightning, and other magnetic interference. The system is certified according to DO-160G and MIL-810H, guaranteeing high resistance to environmental stress. 

Competitive Strengths 

• Engineered to meet aviation and military environmental standards and compatible with a range of UAVs, including larger VTOL and fixed-wing models. 

• George incorporates top-of-time GPS, C2 Radio, IMU, and military-grade sensing technologies, ensuring unprecedented accuracy and situational awareness. 

• Plug-and-play solution with an option to integrate additional avionics modules and different ground infrastructure.

3. Spatial FOG Dual 

Source: Advanced Navigation

Advanced Navigation is a long-standing leader in inertial navigation systems (INS), offering a range of Micro-Electro-Mechanical Systems (MEMS) and Fiber Optic Gyroscopes (FOG) INS/GNSS solutions. FOGs provide greater accuracy in measuring rotational movements and perform better in dynamic conditions (e.g., rapid maneuvers). They also generate less noisy measurements and require less frequent recalibration.

Spatial FOG Dual combines high-accuracy fiber optic gyroscopes, accelerometers, magnetometers, and a pressure sensor with a dual antenna RTK GNSS receiver.  Its EMCORE TAC-450 fiber optic gyro IMU, provides superior inertial data accuracy, surpassing the best MEMS technology.  The RTK GNSS receiver provides positioning accuracy of up to 8mm (0.3 in) and a timing accuracy of 20 ns.  It also supports post-processing kinematics (PPK). 

Similar to Bavovna, Advanced Navigation uses a pre-trained AI neural network sensor fusion algorithm to ensure smooth aerial navigation, even in challenging conditions. The company says its algorithm is 10X more accurate than a traditional Kalman Filter. This allows for extended reckoning, even without GNSS. 

The navigation unit is based on a safety-oriented real-time operating system, with all software designed with high fault tolerance in mind. Hardware is rugged and environmentally protected, in line with IP67 and MIL-STD-810G standards. 

Competitive Features 

• Horizontal position accuracy of 0.01 m with RTK or Kinematica PPK, and velocity accuracy of 0.005 m/s.

• Gyroscopes with a bias instability of 0.1 °/hr and accelerometers with a bias instability of 15 µg. 

• Proprietary, pre-trained neural network sensor fusion algorithm that demonstrated strong results in the field. 

4. Honeywell Compact Inertial Navigation System

Source: Honeywell 

Honeywell Compact Inertial Navigation System (INS) is another great compact option (with a weight of 115 gr/4oz) that packs a punch in terms of hardware. The system includes tactical-grade inertial sensors, dual-antenna GNSS heading, RTK, and closed-loop integration with Pixhawk 2.1 — flexible autopilot. It can be also extended with extra data sources to support specific use cases. For example, you can combine it with a radar velocity system for extra precision or a third-party anti-jamming system for navigating in high EMI environments. 

In an RTK configuration, Honeycomb CINS delivers an ultra-low positioning error of 0.03/0.015 meters (and 2. Meters without RKT). The velocity error is just 0.02-0.04 m/s. It’s a great solution for applications where accuracy is critical.

On the downside, Honeycomb doesn’t use AI for sensor data fusion, unlike Bavovna and Advanced Navigation to provide a greater degree of autonomy and reliability in GPS-compromised environments.

Competitive Strengths:

• Compact INS offers high-accuracy output position, orientation, and velocity data, thanks to top-of-the-range hardware. 

• Redundant sensors and ruggedized system design lend extra safety to flight missions. 

•  Easily extendable with extra Honeycomb or third-party navigation aids systems. e.g. velocity aiding systems

TL: DR: Comparison of the Best INS for UAVs  

Here’s how the top solutions stack in four key areas: customization, SLAM support, non-GPS navigation, and AI sensor fusion. 

	Customizable IMU integrations	SLAM support	Non-GPSnavigation	Custom AI sensor fusion algorithm
Bavovna AI Navigation Kit	✔️	✔️	✔️	✔️
George Autopilot from Avionics	✔️	❌	✔️	❌
Spatial FOG Dual from Advanced Navigation	✔️	❌	✔️	✔️
Honeywell Compact INS	✔️	❌	✔️	❌

Mining is critical for a steady supply of energy resources and rare earth metals for technological advancements. Yet, it’s also a tough business with enduring operating conditions, safety hazards, and stringent regulations.

Many operating processes require aerial visibility – drones have become indispensable for land surveying, risk assessments, stockpile management, and an array of other use cases.

3 Use Cases of Drones in The Mining Industry 

Since 2016, 70% of large mining companies have gone from using drones for fringe cases to fully integrating the technology into surveying, engineering, and monitoring workflows.

Learn how leaders are using drones to optimize operations, cut costs, and increase employee safety.

Streamlined Land Surveying

Mines need land surveying services constantly and in large quantities. Aerial monitoring ensures safe on-site operations, resource assessments, and expansion planning.

Traditional land survey methods like triangulation, tachymetry, chane and tape surveying are time-consuming and require a large on-foot workforce. Surveying personnel at an active mine can interfere with operating equipment, leading to costly downtime. Larger projects like pit-to-port road construction or new mineral exploitation require coverage of large distances. 

Drones are a more time- and cost-effective alternative to conventional land surveying methods. The best fixed-wing and quadcopters with a gas-electric powertrain can stay in the sky for several hours, covering 250-1000+ acres in one flight.

The selection of land-surveying drone payloads available is impressive: HD RGB, oblique, thermal, and multispectral cameras, LiDAR sensors with centimeter-level positioning accuracy, magnetometers, and photogrammetric cameras. Thanks to advances in aerial autonomous navigation, the best drones for mining now have automated data capture modes and can generate 3D environment maps for better situational awareness.

Roughly 83% of mining companies use drones for mapping and surveying, including the Baorixile Open Pit Mine. With a 50 square kilometer mining license area and 640 million tons of coal reserves, the mine is crucial for the local market. However, harsh environmental conditions — snow, sub-zero temperatures, icing — make land surveying missions challenging. Likewise, naturally occurring electromagnetic interference (EMI) hampered drone navigation, causing GPS signal loss and IMU instability. So instead of sending personnel, the operators switched to using Jouav CW-25E UAV with LiDAR technology. 

The drone collected 125 times more data points per square meter than conventional RTK surveying, with under 5 cm accuracy. Data acquisition times were reduced from 13 days to 2 — a 6.5X increase in efficiency. Personnel no longer had to endure the harsh climate, controlling the drone from a parked vehicle.

Stockpile Management

Stockpile management is another labor-intensive task in mining operations. Accurate volume and mass calculations are challenging due to extreme heights and irregular shapes (which also change frequently). However, ensuring structural stability is crucial to prevent accidents. Stockpile collapses and nearly avoided accidents trigger regulatory investigations and fines.

With drones in mining, companies can produce aerial terrain models of their inventory and keep track of stockpile changes over time to ensure safe and effective operations.  Ferrexpo Yeristovo Mine regularly employs drones to conduct land surveys, create up-to-date stockpile maps, and minimize risk to personnel. 

DroneUA helped Yeristovo mine operators create a 3D model of an 82-hectare / 202-acre open pit, collecting data from 410 meters/1345 feet below ground. The team collected over 1,000 data points for 2D/3D models. 

The second major project involved mapping an ore stockpile. The team created pre-planned navigation routes to automate data collection and ensure high data fidelity. Stockpiles occupy large areas with uneven terrain, making hand-held tools less effective. Drones, equipped with photogrammetric cameras, allow greater precision and keep personnel out of dangerous areas. In both cases, drone-led data collection took 90% less time than traditional methods.

Mineral Exploration 

Drones in mining provide a new vantage point for exploring new mineral sites, as well as resource evaluation and subsequent excavation planning. Thanks to advanced onboard sensing technologies and specialized payloads (e.g., an echosounder, metal detector, magnetometer, etc), drones can be used to perform geological mapping missions at faster speeds. For example, identify ore bodies’ spatial distribution or detect tramp metal in stockpiles to prevent costly crusher outages during the excavation phase. 

Yet, drone deployments on mining sites can be challenging due to high EMI levels. Some minerals (e.g.,  magnetite or pyrrhotite), as well as mining equipment like proximity sensors or high-voltage power cables, can interfere with the navigation systems. Bavovna developed an AI-driven hybrid navigation kit for VTOL, Fixed Wing, Multirotor, or Quad to safely cruise in high EMI environments. With Bavovna’s AI navigation kit drones can maintain ​​99% positional accuracy in high EMI/GPS-denied environments and reliably follow the original flight path despite the interferences. The navigation unit is housed in an EMI-protected box, attenuating interferences at 92dB per mm. 

Unlike computer vision-based systems, which require substantial computing resources, Bavovna’s solution relies only on device sensor data, streamed into a lightweight AI model for fully autonomous operations.

Underground tunnel mapping is another labor-intensive task at new mining sites. Irregular tunnel shapes and confined spaces complicate equipment placement and safety. Physical obstructions and geological interferences render traditional surveying methods ineffective. High dust and low lighting limit the usage of optical surveying tools.

Confined space drones have emerged to address these challenges. Featuring a rugged, caged design and advanced obstacle detection, the best indoor drones can cruise through narrow passages and access restricted areas to collect mapping data. Elios 3 from Flyability uses LiDAR and computer vision technologies to generate real-time 3D-environment maps for piloting in GPS-denied environments.

Similar to above-ground land surveying, indoor drones create operational efficiencies and savings. At Glencore Kidd Mine in Ontario operators used Elios 3 to eliminate time-consuming drilling and expensive equipment movement. Within 15 minutes, the team inspects flagged areas and gains video footage and 3D maps of underground conditions.

Conclusion

The usage of drones in mining has grown at a massive clip over the years and for some good reasons. Aerial surveillance capabilities, coupled with a great array of payloads, help miners conduct faster site assessments, survey underground tunnels, and improve asset management lifecycles. 

Thanks to advances in onboard sensing capabilities and new navigation technologies like Bavovna’s kit, drone flights have also become much safer even in GNSS-denied and tough environmental conditions, paving the way for even more deployment scenarios. 

Global Navigation Satellite Systems (GNSS) systems such as GPS have been a staple technology in land, maritime, and aerial navigation for decades. Super useful? Absolutely. But, GNSS has its fair share of downsides too.

Weather, buildings, natural obstacles, and some electronic systems cause inevitable signal degradation and latency in signal acquisition. Heavy reliance on a network of satellites and ground stations, attractive targets for physical or cyber-attacks, also raises rightful security concerns. For UAVs and ROVs, signal spoofing, jamming, and other forms of electromagnetic warfare (EW) are also a problem. 

For many mobility scenarios signal loss or high latency is unacceptable, even more so if we’re to advance to full autonomy. But if not GNSS, what other technologies can provide reliable positioning and navigation data? A couple of options are on the table. 

Sensor Fusion with AI 

Every commercial UAV is equipped with a set of standard sensors, such as an accelerometer, gyroscope, compass, airflow sensor, and altimeter, for situational awareness. More advanced enterprise models may also include LiDAR distance sensors or ultrasonic sensors to detect proximity to obstacles.

Combined, these provide real-time information on the drone’s location, orientation, and movement, and a GPS/GNSS receiver tallies this with geo coordinates, which allow the drone to follow a pre-planned flight route. When GPS is down, the best models will hold position or attempt to automatically return home. 

Yet, recent advances in sensor fusion also enable autonomous navigation scenarios. Instead of the GPS signal and pilot commands, the drone can rely on data from its Inertial Measurement Unit (IMU) and AI algorithms to safely continue the route. By integrating and processing sensor data streams on the device, AI pilot models can provide precise situational awareness to drones and full autonomy. 

Bavovna Navigation Kit includes an 800gr/28 oz onboard edge devices and SaaS access to a fine-tuned AI Hybrid navigation model, pre-trained on historical data from the available sensors. Already implemented on eight UAV platforms, ranging from fixed-wing and VTOLs to multi-copters and FPVs, the system maintains an End Point Positioning Error (EPPE) of under 0.5% on complex trajectories of up to 30 km or 18.6 miles. 

Intended for dual use, Bavovna’s hardware sits in an EMI-protected case and can be further augmented with a SIGINT RF module for bypassing EW and EM obstacles. AI pilot system, not only enables more reliable autonomous navigation but also improves obstacle detection, UAV routing, and ultimately operating efficiencies. 

LiDAR-Based SLAM Systems

An alternative approach to sensor fusion is the Simultaneous Localization and Mapping (SLAM) technique. While sensor fusion primarily relies on IMU data, SLAM algorithms use camera and LiDAR data to general a real-time map and localize the piloted system in a GPS-denied environment. Opposed to cameras, 3D LiDARS provides an immediate point cloud of the environment, eliminating the latency. Data processing software has to be installed both on the flight controller and the onboard computer. 

SLAM algorithms like LOAM and Cartographer have demonstrated great performance during scientific experiments in terms of trajectory execution and landing errors. Commercially, Elios 3 from Flyability and Hovermap series from Emesent, and Scout 137 from ScoutDi rely on SLAM for indoor navigation in GNSS-denied environments (with an added bonus of advanced mapping scenarios). 

The downside of photo-SLAM, however, is high resource consumption: CPU, memory, and processing, especially for photo-realistic mapping scenarios. This, in turn, requires heavier and more expensive onboard units, which increase the system operating costs. Likewise, high resource consumption results in faster battery drainage on smaller UAVs, undercutting the precious flight time. Although some recent advancements in processing frameworks help optimize resource usage. 

AI Pilots: Navigation Guided by Magnetic Fields 

Commercial and military aircraft are also increasingly relying on AI-assisted navigation, oftentimes powered by hybrid INS/GPS signals. For security reasons, however, the military is looking to progressively minimize reliance on satellite networks.  

The US Air Force has been continuously testing different AI pilots to autonomously pilot fighter jets. Technologically, we’re already there — AI systems successfully piloted an F-16 in a dogfighting exercise over California’s Edwards Air Force Base in May 2024. But there’s always room for further improvements. 

The Air Force research team is working on a new AI navigation pilot to potentially navigate the plane using the Earth’s magnetic fields. So far, such scenarios haven’t been feasible as the aircraft generate too much electromagnetic noise, interfering with onboard magnetometer measurements.

However, a recent model, tested on C-17, showed very promising results that could lead to “tactical airdrop quality” and a potential solution to “things we can do, should we end up operating in a GPS-denied environment.“, according to Col. Garry Floyd, director of the Department of Air Force-MIT Artificial Intelligence Accelerator program.

While we may not be there yet, widespread autonomous AI piloting without GNSS is no longer a question of “if”, but rather “when”, especially for unmanned aerial vehicles.