From Steep Slopes and Glaciers to High Speed: Next-Generation Autonomous Drone Landings

What if you were working in an environment with really steep gradient, and you had to land your drone without risking it slipping down the steep slopes?

Because, let’s face it, a drone is a really expensive investment, and the one thing you do not want is to lose your investment because you risked landing it onto a slippery surface, only to watched it hurtle down into mortal contact.

But we could now be heading somewhere though – researchers at the University of Sherbrooke’s 3IIT Institute in Canada have asked that question and returned with a solution they figure is comfortable landing on any surface, steep or flat.

The team led by Professor Alexis Lussier Desbiens — who directs research across aerial robotics and sports engineering — is tackling this challenge through integrated hardware and software co-design.

Working alongside him are post-doctoral researcher Isaac Tunney, who completed his PhD in mechanical engineering before specialising in robotics, and doctoral candidate John P. Bass, who is nearing completion of his own mechanical engineering PhD.

The 3IIT institute plays a pivotal role in the region’s technology sector, serving as an industrial bridge between fundamental academic research and commercial prototyping through short-term contracts and five-year partnerships with industrial clients.

Located four hours east of the university’s main campus, the institute operates within uncontrolled airspace, granting researchers the rare ability to fly and iterate experimental aircraft directly on site.

The team’s solution is centred around microspines – one of many ways to enable robots to latch onto surfaces like walls and ceilings; with the Canadian roboticists using the mini spikes to get drones to land on more challenging, remote surfaces such as icebergs.

Called the Ice Dart, the drone is like a spider, and can land and latch onto steep, slippery surfaces such as icebergs and glaciers — an increasingly useful capability as activity in the Arctic increases. The drone can grip onto icy slopes of nearly 60 degrees, which is way beyond what most humans could manage without special equipment.

In the study article below, the researchers explained how they developed the Ice Dart drone with a special landing gear that absorbs the impact of a hard landing while holding the drone in place with tiny spines that penetrate and grip the ice.

Published in IEEE Transactions on Field Robotics, the study describes how the Ice Dart was able to land on icebergs and a glacier in southeast Iceland.

Tests took place amid persistent winds and temperatures of 0 to 10°C along the ruggedly breathtaking Fjallsjökull (pronounced “FYATLS-yuh-kuutl”) glacier, which empties into a lagoon filled with icebergs.

The drone was able to successfully perch at speeds of up to three metres per second and slopes of up to 58 degrees, with a success rate of 100 percent even in wind speeds of 30 km/h.

The researchers were motivated by a desire to allow drones to land almost anywhere in the world, since the availability of safe landing sites is one of the primary limitations on where and how drones can operate. The researchers already have a history of developing drones that can land on fast-moving trucks as well as trailers, boats, and steep roofs.

“The ability to land rather than hover can fundamentally change how drones are used in the field,” says Desbiens.

“Once a drone has landed, energy consumption drops dramatically, allowing much longer observation periods with a small aircraft. The drone also becomes completely silent and can even reduce or eliminate its thermal and RF signature by shutting down major onboard systems.”

Landing on icebergs specifically allows drones to monitor them for days or months, producing more detailed observation than a quick aerial surveillance mission.

This could simplify iceberg tracking compared to methods such as helicopter deployment, dropped instruments, or dart-like tracking devices, and provide another data layer to satellite and ship-based iceberg detection, according to the researchers. It could also provide a means of monitoring icebergs that are otherwise untrackable.

With its carbon-fibre construction, the Ice Dart drone weighs just 2.65 kg and has four legs arranged in an X shape, attached to its body with a pivot joint. Used in the group’s previous drone research, this landing gear disperses energy to reduce impact and overcomes multiple engineering challenges.

The friction shock absorbers consist of 38 disks that generate friction torque as the legs move up and down upon impact. This lowers the UAV’s centre of mass and helps spread out the kinetic energy of landing, but the real trick comes in the form of two retractable spines on each foot—one for uphill and one for downhill grip.

The larger spine engages on the more heavily loaded downhill feet, and the smaller, thinner spine engages more easily on the uphill feet, even under very low loads on steep slopes. The spines only penetrate the ice as the suspension compresses, generating grip and protecting them from high-impact forces.

“The inspiration for the retractable spines in the feet came from looking at a cat’s claws and their ability to deploy only when needed,” says Tunney.

“I wanted to create feet that would naturally and passively engage their spines in the ice at the right moment, regardless of the drone’s orientation, the surface geometry, or the ice conditions.”

Now, for those who want details, the minute and technical details are shared in the below article.

UAS have matured into indispensable instruments across environmental monitoring, industrial inspection and emergency response.

However, their operational utility remains constrained by a persistent physical bottleneck: the landing envelope. While advances in sensor payloads, onboard processing and flight control algorithms allow aircraft to navigate complex airspaces with high precision, touchdown remains largely restricted to flat, benign surfaces.

Expanding this operational envelope to encompass steep inclines, fast-moving target platforms and low-friction ice formations is critical for unlocking the full potential of autonomous aerial systems.

This flight ecosystem is backed by motor test benches, fabrication laboratories, a closed-loop wind tunnel, an indoor flight room equipped with a wind generator and access to a local airport runway 30 minutes away that can be closed on demand for high-speed trials.

Through this infrastructure, the team evaluated the fundamental mechanical limitations of conventional landing systems. Traditional rigid landing gear acts as an undamped spring, forcing aircraft to maintain ultra-slow descent rates of approximately 0.5 m/s.

These prolonged descents expose drones to atmospheric wind gusts, accentuate aerodynamic disturbances and increase the risk of sensor degradation. When impact speeds increase, rigid structures generate violent rebounds and extreme impact forces that destabilise the craft.

This problem is compounded by standard flight controllers, which delay motor thrust cutoff until state stability is verified; during this latency window, reoriented residual thrust actively destabilises the craft and flips it off the landing surface.

Overcoming this traditional 0.5 m/s descent constraint fundamentally transforms drone utility, shifting UAVs from fragile laboratory instruments into robust, all-weather field tools capable of aggressive touchdowns.

Eliminating this mechanical vulnerability requires a paired intervention: mechanical energy dissipation during impact and active software control to secure surface adhesion on steep slopes.

Steep-slope and Rooftop Landings: Friction Dissipation and Reverse Thrust

Landing an aerial robot on steep slopes inclined up to 60° introduces severe rotational and translational instability. At touchdown, kinetic energy must be absorbed immediately to prevent the aircraft from rebounding down the incline, while mechanical friction or downward force must be generated to prevent gravity from pulling the craft downhill.

Standard rigid gear amplifies these forces, causing continuous rebounds or inducing catastrophic forward flips as kinetic energy transfers into pitch rotation.

To resolve these dynamics, the 3IIT team engineered a dissipative suspension system utilizing a stacked friction disc shock absorber mechanism.

An integrated ten-gram DC motor adjusts the compression force applied to the disc stack, allowing researchers to tune the energy dissipation rate to match expected impact velocities without requiring complex, heavy real-time active control feedback loops during impact. Internal torsion springs reset the leg geometry once joint friction is relieved.

Complementing this hardware, the researchers modified the ArduPilot landing mode software to enable motor reverse thrust upon contact detection. When onboard inertial measurement unit (IMU) accelerometers register impact, the flight controller scales motor PWM output across a 1500 to 2000 range within a custom landing mode flag, generating reverse thrust that actively pushes the aircraft against the incline.

The mechanical suspension and control software function synergistically: the dissipative friction joints stretch the total impact duration, creating the required time window for the reverse thrust command to spool up and generate firm surface adhesion.

In field tests, this dual strategy enabled reliable touchdowns on slopes up to 60° at impact descent speeds reaching 2.75 metres per second (m/s) — representing an eightfold expansion of the landing envelope compared with conventional rigid gear.

This automated capability allows drones to land, reconfigure leg geometry and re-take off autonomously across sloped rooftops in North America, establishing a foundation for repeatable urban deployment and emergency touchdowns.

However, transferring these kinetic absorption principles from stationary slopes to fast-moving targets introduces severe aerodynamic drag and tracking latency demands.

High-speed Vehicle Integration: The Dart UAV and Ultra-low Latency Telemetry

Landing an autonomous drone on a high-speed ground vehicle presents complex aerodynamic and telemetry challenges.

At high forward velocities, the drone must maintain a steep forward pitch angle to overcome drag, creating a substantial risk that the propellers or landing gear will strike the vehicle surface unevenly upon descent.

Furthermore, descending through the turbulent airflow boundary layer generated by the target vehicle induces severe relative velocity fluctuations. If an aircraft attempts to level its attitude too early, it suffers immediate relative velocity loss and misses the target; if it levels too late or with excessive angular momentum, the impact dynamics inevitably precipitate a rollover crash.

To survive these high-impact dynamics, the team developed a specialized aircraft designated the Dart UAV (Direct Approach Rapid Touchdown). Engineered with a thrust-to-weight ratio of seven and capable of airspeeds up to 126 km/h, the Dart UAV features high-impact structural damping designed to dissipate impacts between 4m/s and 6m/s — equivalent to a freefall from a height of two metres.

The entire landing gear accounts for fourteen percent of the aircraft’s total mass, with the friction joints comprising approximately five percent.

To execute precise target tracking, a target beacon equipped with two GPS units (CubePilot Cube+ hardware) is mounted to the vehicle, providing accurate yaw alignment.

The underlying communications architecture splits telemetry tasks to balance accuracy and speed: standard RTCM RTK differential positioning corrections are transmitted at 1 Hz over an RFD900s radio link from the ground station to the drone and relayed to the target beacon.

Meanwhile, high-rate target state data — encompassing position, velocity and heading — are streamed directly between the target beacon and the drone at 20 Hz using ESP32 radio modules running the ESP-NOW protocol. Compared to baseline RFD900s communications, which suffered from flaky 5 Hz updates and latencies between 100ms and 160ms, elevating MAVLink task priorities and increasing UART baud rates within custom ArduPilot code allowed the team to achieve a ultra-low telemetry latency of 25ms.

Combined with algorithmic latency compensation that projects target location based on vehicle velocity, the Dart UAV executes a structured two-stage landing sequence:

1. Target Tracking and Rapid Descent: The drone tracks the target beacon at a predefined height and executes a rapid vertical descent to punch through the turbulent boundary layer above the vehicle.
2. Pre-impact Leveling and Touchdown: At a precise altitude above the target, the control system executes a rapid leveling maneuver, commanding pitch and roll to zero before impact. Upon IMU impact detection, reverse thrust engages to lock the drone onto the surface.

Across 38 drag strip field trials conducted on a rented runway at vehicle speeds between 10km/h and 110km/h and descent speeds of 3m/s, the system achieved a 100 per cent landing success rate without surface sliding.

Secondary trials successfully validated landings on dirt roads, moving trailers and winch-pulled canoes, while software-in-the-loop (SITL) simulations demonstrated a 60-fold landing envelope expansion over baseline hardware. An additional trial at 120 km/h revealed the physical boundary condition of the system: while the aircraft successfully landed on the moving vehicle, the extreme momentum induced significant surface slipping, defining the operational velocity limit of the friction-based deceleration model.

These low-latency control architectures enable reliable autonomous collaborative fleet operations across moving land assets and sea vessels.

Cryospheric Missions: Deploying Passive Retractable Spine Footing in Iceland

Deploying aerial robots in cryospheric environments presents severe physical challenges: ice surfaces offer near-zero friction, ambient temperatures degrade battery output and extreme wind gusts threaten total asset loss.

Standard friction-based landing gear fails on smooth ice, where residual motor thrust or lateral wind forces cause the aircraft to slide uncontrollably into surface crevasses, meltwater puddles or structural protrusions.

To address these low-friction conditions, the researchers engineered an ice-landing variant of the aircraft. The structural design features arched carbon-fibre leg geometry to clear surface bumps and terrain irregularities, integrated flotation foam to guarantee buoyancy and asset recovery in open water, and weatherproofing against rain and snow.

Although the added mechanical mass restricts flight endurance to approximately ten minutes, this operational window is sufficient for targeting local ice formations.

The core innovation rests in the foot mechanism: passive opposing retractable spines designed to deliver aggressive directional braking regardless of landing orientation.

As the drone makes contact and slides along an ice surface, the single spine aligned with the vector of travel naturally engages and digs into the ice, while opposing spines passively retract into the foot structure.

Extensive load research dictated exact spine sizing to maximize mechanical grip without overloading the leg joints or requiring motorized actuators and complex control loops.

During a field campaign in Iceland, the team deployed three aircraft to conduct surface touchdowns across 22 icebergs and two glaciers.

The operational workflow relied on coordinated teamwork: one researcher (John) piloted a scouting drone to locate viable iceberg landing zones, while the second researcher (Isaac) executed the automated descent and reverse-thrust sequence on the target craft via a single-switch pilot command.

The drones completed 24 successful ice landings on slopes inclined up to 60°, and all three aircraft were recovered intact without a single crash. Embodying passive mechanical intelligence within the footing mechanism eliminates active actuator loops, maximising system reliability under harsh polar conditions.

Strategic Implications for the Unmanned Aerial Vehicle Sector

The research conducted at the 3IIT institute demonstrates that hardware-software co-design can successfully eliminate touchdown dynamics as the central bottleneck in UAV operations. By unifying dissipative mechanics with responsive flight control software, the team established three distinct technical approaches:

• Stacked friction disc suspensions that stretch impact duration and absorb high descent kinetic energy.
• Ultra-low latency ESP32 telemetry that enables real-time tracking and pre-impact leveling on fast-moving targets.
• Passive retractable spine footing that delivers directional mechanical braking on zero-friction ice surfaces.

Looking ahead, future development vectors include integrating real-time computer vision algorithms for precise feature tracking and incorporating load-cell test arrays to map impact force frequencies and acceleration profiles.

Key Takeaways

• Operational Shift: Moves aerial robotics focus from airborne flight optimisation to landed-state mission execution, conserving battery power while maintaining persistent presence.
• Environmental Surveillance: Enables long-term monitoring, climate data collection and cliffside botanical sampling of endangered plant species (as demonstrated by laboratory spin-off Outreach Robotics).
• Industrial Asset Inspection: Supports physical contact monitoring on high-voltage power lines (in collaboration with Hydro-Québec) and autonomous mapping inside underground mines using tethered LiDAR systems.

By expanding the physical operational envelope to accommodate steep slopes, high-speed ground targets and slippery glacial ice, this integrated engineering approach alters how industry, science and defence sectors structure remote autonomous deployments worldwide.