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.