高速ドローンがGPSなしで森林を航行

Researchers from the HKU Mechatronics and Robotic Systems Lab (MaRS LAB) have introduced a navigation system that enables high-speed drones to detect centimeter-thick items while in rapid flight. The system combines advanced sensors, lightweight drone components, and a new flight path framework to dodge obstacles in real-time while speeding through congested areas. Here’s what you need to know.

高速ドローンが鳥の飛行を模倣するように設計されている方法

There is no denying that drone technology has undergone significant improvements over the last few years. Today’s drones can travel further, carry more weight, and support more sensors than their predecessors. Additionally, they are much cheaper to produce and more readily available.
Despite these advancements, there is yet to be a drone that can match the capability of a bird in terms of determining flight paths in real time while accounting for unexpected variables like branches or other birds. Drones can travel long distances and land on precise locations. However, these devices still rely on pre-determined flight paths or external sensory input like GPS to navigate to their targets.

自律的な高速ナビゲーションが課題となってきた理由

To date, researchers haven’t been able to figure out how to achieve autonomous, safe, and high-speed drone navigation without relying on external guidance or pre-mapped routes. The problems derive from several issues, including the need to add more sensors in order to enable these crafts to detect more obstacles.

As you introduce more sensors, additional computing power is required to process the added sensory input data. The stronger the computer and sensors needed, the more weight that gets added to the drone, reducing its ability to fly fast and remain stable at speed.

Thankfully, these restrictions may have been overcome by a team of ingenuitive minds from the Department of Mechanical Engineering of the Faculty of Engineering at the University of Hong Kong (HKU).

研究概要：高速ドローンの安全保証ナビゲーション

Scientist Yunfan Ren and a team of engineers published¹ the Safety-assured high-speed navigation for MAVs in Science Robotics recently. The study sheds light on a next-generation autonomous flight technology designed to enable drones and micro air vehicles (MAVs) to navigate complex environments at top speeds without external sensory input.

SUPERドローンとは？仕様と特徴

The SUPER drone is a compact MAV purpose-built to demonstrate the engineer’s new flight path system. The device was designed from day one to be light and agile. It features a compact form factor with a 280-millimeter wheelbase and micro computer elements that give it an impressive thrust-to-weight ratio of +5.0 and a takeoff weight of 1.5 kg. These factors enable the SUPER to travel at +20 meters per second while dodging obstacles.

出典 – 香港大学

高速性能のための軽量コンピューティング

At the core of the SUPER is a miniature computer that can monitor and compute sensory input. The lightweight board provides the craft with more flight time while improving its capabilities compared to its predecessors. Additionally, the flight computer can support waypoints, GPS, and other forms of external navigation when needed.

LiDARとセンサーが自律飛行を導く仕組み

The onboard sensor suite is what enables the craft to operate with such high performance and under different conditions. The unit incorporates a lightweight 3D light detection and ranging (LIDAR) sensor that enables it to determine items as thin as a twig from +70 meters away. Notably, the LiDAR input is entered directly as point clouds along the route. This approach provides much faster input and response times for the flight path computer.

リアルタイムの飛行経路計画と調整

The engineers took a unique approach to flight planning. For example, the system constantly scans and updates flight paths for the craft. The new flight path coordinator creates two different paths as options for the drone. The first path is the clearest route, while the second path has unknown vectors. From there, the computer utilizes real-time data input to adjust the patterns based on the situation.

実環境での高速ドローンナビゲーションテスト

The engineers conducted several tests to ensure that their high-speed drones would operate as predicted. They conducted multiple flight tests where the craft had to navigate complex terrain. Additionally, the engineers utilized different lighting conditions, including pitch black, to see how the sensor’s performance was affected.
Interestingly, the team selected several locations for the testing phase. The testing sites ranged from inside office spaces to outside in dense forests. In one test, the device was sent to whizz through a forest at night. To accomplish this task, the device had to scan, register, and navigate obstacles in real time while traveling at a high velocity.

高速ドローン試験の主要結果

The high-speed drone achieved autonomous flights in several scenarios. It was able to achieve speeds of +20 meters per second while simultaneously navigating intricate environments. While not perfect, the system did register a 35x reduction in flight failure rates using the new method. Additionally, the flight path took half as much time to set up, and the device flew faster than traditional drones in the same testing scenarios.

先進的な高速ドローン技術の利点

There are many benefits to high-speed drones that can navigate autonomously through dense airspace. For one, they can be used in applications where agility and safe navigation are critical to mission success. These devices can make split-second decisions that allow them to determine the best way to dodge an approaching object and remain on course.

搭載インテリジェンスでドローンプログラミングを簡素化

Determining a safe flight path for drones has always been an issue. As more air traffic emerges and drone routes become more complicated, creating flight paths will require even more effort. The autonomous systems introduced in the study could help to reduce these times as the craft would be capable of making flight path alterations when needed in real time.

高速ドローンの実世界での応用とタイムライン

There is a long list of applications for agile autonomous MAVs (Miniature Aerial Vehicles). These devices will open the door for a new level of convenience and response. From emergency crews using these devices to find survivors to e-commerce sites delivering your packages, there are many uses for these high-tech devices.
This high-speed drone technology could be put into use within the next year or less. Several countries are pushing to become global players in the drone market. This technology is sure to be met with open arms as engineers seek to make drones smarter and more aware. You should expect to see military integration within the next 2 years, with commercial allocations hitting the market in the next 5 years.

捜索救助ミッションにおける自律ドローン

One of the key uses for these autonomous craft will be assisting in search and rescue missions. These devices will be able to autonomously explore wreckage following a disaster and locate human survivors. They can call in for supplies and other much-needed help, which can also be delivered via drone to the survivor’s location until human help arrives. Eventually, the concept could scale up to drones capable to carry survivors to safety directly.

高速ドローンがドローン配送を革命的に変える方法

Companies like Amazon have teased drone delivery for years. However, the concept has proven to be much more difficult than previously thought. This latest development should help to push this tech forward as the devices would be able to adjust their flight path to optimize performance. Consequently, next-day delivery may become the norm.

インフラ監視に自律ドローンを活用する

Keeping bridges, roads, and other vital infrastructure in good working condition is a monumental task that requires constant supervision. The use of autonomous drones could help to reduce the manpower needed, while improving the capabilities of those tasked with monitoring these essential items. In the future, drones will scan key infrastructure and notify inspectors of potential dangers before they become a problem.

自律ドローンによる環境モニタリング

Another area where autonomous drones could make a big difference is in environmental monitoring. These systems can be set up to automatically run routes and register any changes in the environment. This approach provides a new level of awareness to environmentalists and could also help scientists keep tabs on the planet’s health.

探索とデータ収集のためのドローンスウォーム

There are many who see autonomous drone swarms as one of the best ways to discover and map out new areas. This approach eliminates the risks to explorers and provides real-time data for researchers to review and determine the next best course of action. Already, autonomous drones have proven crucial in the mapping industry, where they are used to track hard-to-reach locations like the ocean floor.

高速ドローン研究者

Yunfan Ren was the lead author of the high-speed drone research paper. He had support from Professor Fu Zhang and a team of researchers from the Department of Mechanical Engineering of the Faculty of Engineering at the University of Hong Kong (HKU).

高速ドローンの将来開発目標

The race is on to improve SUPER’s range, speed, and processing power. The team will now seek to push the boundaries of SUPER to see how to make the craft lighter, more agile, with longer flight times, and capable of heavier payloads. Also, the engineers seek to miniaturize the MAV further, enabling it to be used in more scenarios.

ドローンセクターへの投資

The drone market is a fast-growing sector that has competition from across the globe. Notably, China’s DJI holds the lion’s share of the drone market. However, many firms have the potential to upend the industry with their new products and business models. Here’s one company positioned for success.

EHang 

Ehang (EH ) entered the drone industry in 2014 to make passenger-carrying drones a reality. The company remains pivotal in taking drones from tools to a form of reliable transport. To that end, it has achieved several accomplishments, including launching the world’s first person-carrying Autonomous Aerial Vehicle (AAV).

In 2018, Ehang signed several high-end partnerships and development deals. Notably, one of these agreements included working with the city of Lyon, France,  on an air taxi protocol. Another partnership, with the Austrian aerospace group FACC,  sought to improve passenger drone capabilities and safety.