Chris Anderson's Posts (2716)

Sort by
3D Robotics

Microsoft AirSim drone racing at NeuroIPS

5334386873?profile=original

From the Microsoft Research Team:

Drone racing has transformed from a niche activity sparked by enthusiastic hobbyists to an internationally televised sport. In parallel, computer vision and machine learning are making rapid progress, along with advances in agile trajectory planning, control, and state estimation for quadcopters. These advances enable increased autonomy and reliability for drones. More recently, the unmanned aerial vehicle (UAV) research community has begun to tackle the drone-racing problem. This has given rise to competitions, with the goal of beating human performance in drone racing.

At the thirty-third Conference on Neural Information Processing Systems (NeurIPS 2019), the AirSim research team is working together with Stanford University and University of Zurich to further democratize drone-racing research by hosting a simulation-based competition, Game of Drones. We are hosting the competition on Microsoft AirSim, our Unreal Engine-based simulator for multirotors. The competition focuses on trajectory planning and control, computer vision, and opponent drone avoidance. This is achieved via three tiers:

  • Tier 1  Planning only: The participant’s drone races tête-à-tête with a Microsoft Research opponent racer. The goal is to go through all gates in the minimum possible time, without hitting the opponent drone. Ground truth for gate poses, the opponent drone pose, and the participant drone are provided. These are accessible via our application-programming interfaces (APIs). The opponent racer follows a minimum jerk trajectory, which goes through randomized waypoints selected in each gate’s cross section.
  • Tier 2  Perception only: This is a time trial format where the participants are provided with noisy gate poses. There’s no opponent drone. The next gate will not always be in view, but the noisy pose returned by our API will steer the drone roughly in the right direction, after which vision-based control would be necessary.
  • Tier 3 – Perception and Planning: This combines Tier 1 and 2. Given the ground truth state estimate for participant drone and noisy estimate for gates, the goal is to race against the opponent racer without colliding with it.

The animation on the left below shows the ground truth gate poses (Tier 1), while the animation on the right shows the noisy gate poses (Tier 2 and Tier 3). In each animation, the drone is tracking a minimum jerk trajectory using one of our competition APIs.

Image shows the ground truth gate poses

 

The following animation shows a segment of one of our racing tracks with two drones racing against each other. Here “drone_2” (pink spline) is the opponent racer going through randomized waypoints in each gate cross section, while “drone_1” (yellow spline) is a representative competitor going through the gate centers.

This animation shows a segment of one of our racing tracks with two drones racing against each other

The competition is being run in two stages—an initial qualification round and a final round. A set of training binaries with configurable racetracks was made available to the participants initially, for prototyping and verification of algorithms on arbitrary racetracks. In the qualification stage (Oct 15th to Nov 21st), teams were asked to submit their entries for a subset or all of the three competition tiers.  117 teams registered for the competition worldwide, with 16 unique entries that have shown up on the qualification leaderboard.

We are now running the final round of the competition and the corresponding leaderboard is available here. All of the information for the competition is available at our GitHub repository, along with the training, qualification, and final race environments.

Engineering-wise, we introduced some new APIs in AirSim specifically for the competition, and we’re continually adding more features as we get feedback. We highlight the main components below:

In the long term, we intend to keep the competition open, and we will be adding more racing environments after NeurIPS 2019. While the first iteration brought an array of new features to AirSim, there are still many essential ingredients for trustable autonomy in real-world scenarios and effective simulation-to-reality transfer of learned policies. These include reliable state estimation; camera sensor models and motion blur; robustness to environmental conditions like weather, brightness, and diversity in texture and shape of the drone racing gates; and robustness against dynamics of the quadcopter. Over the next iterations, we aim to extend the competition to focus on these components of autonomy as well.

For more of the exciting work Microsoft is doing with AirSim, see our blog post on Ignite 2019.

Read more…
3D Robotics

Evolution of solar-powered drones

From Hackaday:

Many of us have projects that end up spanning multiple years and multiple iterations, and gets revisited every time inspiration strikes and you’ve forgotten just how much work and frustration the previous round was. For [Daniel Riley] AKA [rctestflight] that project is a solar powered RC plane which to date spans 4 years, 4 versions and 13 videos. It is a treasure trove of information collected through hard experience, covering carbon fibre construction techniques, solar power management and the challenges of testing in the real world, among others.

Solar Plane V1 had a 9.5 ft / 2.9 m carbon fibre skeleton wing, covered with transparent film, with the fragile monocrystaline solar cells mounted inside the wing. V1 experienced multiple crashes which shattered all the solar cells, until [Daniel] discovered that the wing flexed under aileron input. It also did not have any form of solar charge control. V2 added a second wing spar to a slightly longer 9.83 ft / 3 m wing, which allowed for more solar cells.

Solar Plane V3 was upgraded to use a single hexagonal spar to save weight while still keeping stiff, and the solar cells were more durable and efficient. [Daniel] did a lot of testing to find an optimal solar charging set-up and found that using the solar array to charge the batteries directly in a well-balanced system actually works equally well or better than an MPPT charge controller.

V4 is a departure from the complicated carbon fibre design, and uses a simple foam board flying wing with a stepped KF airfoil instead. The craft is much smaller with only a 6 ft / 1.83 m wingspan. It performed exceptionally well, keeping the battery fully charged during the entire flight, which unfortunately ended in a crash after adjusting the autopilot. [Daniel] suspects the main reasons for the improved performance are higher quality solar panels and the fact that there is no longer film covering the cells.

Read more…
3D Robotics

From Hackaday:

There’s nothing quite like the sight of a plastic box merrily sailing its way around a lake to symbolise how easy it is to get started in autonomous robotics. This isn’t a project we’re writing about because of technical excellence, but purely because watching an autonomous tupperware box navigate a lake by itself is surprisingly compelling viewing. The reason that [rctestflight] built the vessel was to test out the capabilities of ArduRover. ArduRover is, of course, a flavour of the extremely popular open source ArduPilot, and in this case is running on a Pixhawk.

The hardware itself is deliberately as simple as possible: two small motors with RC car ESCs, a GPS, some power management and a telemetry module are all it takes. The telemetry module allows the course/mission to be updated on the fly, as well as sending diagnostic data back home. Initially, this setup performed poorly; low GPS accuracy combined with a high frequency control loop piloting a device with little inertia lead to a very erratic path. But after applying some filtering to the GPS this improved significantly.

Despite the simplicity of the setup, it wasn’t immune to flaws. Seaweed in the prop was a cause of some stressful viewing, not to mention the lack of power required to sail against the wind. After these problems caused the boat to drift off course past a nearby pontoon, public sightings ranged from an illegal police drone to a dog with lights on its head.

If you want to use your autonomous boat for other purposes than scaring the public, we’ve written about vessels that have been used to map the depth of the sea bedtrack aircraft, and even cross the Atlantic.

Read more…
3D Robotics

Swarming Solos

3689741241?profile=original

Collaboration of Rajant, a mesh networking company, and the Norwegian military. BBC story here;

Scientists from the Norwegian Defence Research Establishment (FFI) and the US's Rajant Corporation are working on simultaneously flying about 20 drones that can work in co-ordination with little human supervision.

A Rajant-patented radio technology called "kinetic mesh" and "foreign function interface" distributed computing software are the technological ingredients behind this breakthrough.

3689741171?profile=original

Read more…
3D Robotics

PX4-based tailsitter VTOL

From Hackster:

"Engineers at the University of Toronto have designed a fully open sourced dual-rotor tail-sitter MAV using readily available electronics and 3D-printed parts. The Phoenix drone is based on the PX4 autopilot platform, PX4 middleware, and is equipped with a Pixracer flight computer, supporting both flight control and ArduPilot’s SITL simulation.

“Our open source package, available on GitHub, includes mechanical design documents, component lists, a carefully tuned and verified dynamics model, control software, and a full set of simulation tools — in short, everything necessary to understand, construct, test and verify a prototypical tail-sitter MAV.”

On the hardware end, the Phoenix is outfitted with a flight computer that packs an STM32F427VIT6 SoC loaded with a Cortex-M4F microprocessor (256Kb of SRAM), and a series of sensors — an Invensense ICM20608 (accel/gyro), an MPU-9250 (accel/gyro/mag), a Measurement Specialties MS5611 barometer, and a HMC5983 magnetometer (with temp compensation).

1*vSGgXs8_Cc-psG5J-oOOlQ.png

The drone’s frame was constructed out of a cast polyurethane foam core and 3D-printed plastic parts, making it extremely light. Driving the Phoenix are a pair of Gemfan 8-inch diameter 4.5-inch pitch propellers, powered by TMotor 2208–18 1100 Kv brushless DC motors and a 2200mAh Li-Po battery.

On the software side, the engineers tasked custom flight-control software based on the Pixracer autopilot platform with PX4 support, along with ESC firmware (based on BLHeli), and the MAVROS robot OS, which they use to tie in the Phoenix to a ground station for control. They also included a MATLABSimulink system and SITL Gazebo to compile and test flight code on a desktop PC. The team states that the Phoenix is a great learning and research platform, and hope educators, hobbyists, and researchers use it to create “innovative new modifications and derivatives.”

Read more…
3D Robotics

From the Next Web:

IBM‘s Developer Drone Drop 2019 contest is officially underway. Now through June 16 the company will give away 1,500 drones to developers who enter. Why is IBM giving away free drones? It hopes you’ll use them to deliver AI-powered solutions to the problems caused by natural disasters.

The contest officially started last week, but there’s plenty of time to sign up. You don’t have to be an expert or have any code built already to enter – the winners will be selected at random, not by judges. Winners will receive more than just a robot, according to IBM:

The DJI Tello drone is more than just a cool prize. We’ll give you code patterns to unlock its potential, and introduce you to new skills around visual recognition, AI and machine learning.

The giveaway comes courtesy of IBM‘s Code and Response, a new initiative this year from the company that aims to empower developers with the resources and support to implement original technology-based solutions to humanity’s open problems. Inspiring developers who, otherwise, might not have access to IBM‘s resources and mentors is a strategy that’s already paying off for the company.

TNW spoke to IBM Code and Response CTO Daniel Krook to ask why IBM was giving away drones for the second year in a row. He told us:

Who doesn’t like free drones? It’s about inspiring people … it’s not just altruism on IBM‘s part, we believe this technology can help humanity and IBM is a part of humanity.

One of last year’s hackathon winners, Pedro Cruz, developed his drone-based disaster relief tech after experiencing the devastation of Puerto Rico by hurricanes Irma and Maria.

Read more…
3D Robotics

Flying stick bi-copter

From Hackaday:

Fixed-wing planes and helicopters are no longer the darling of the RC world. Even quadcopters and other multirotors are starting to look old hat, as the community looks to ever more outrageous designs. [rctestflight] has slimmed things down to the extreme with this coaxial bicopter build, also known as the Flying Stick (Youtube video, embedded below).

The initial design consists of two brushless outrunner motors fitted with props, rotating in opposite directions to cancel out their respective torques. Each is mounted on a gimbal, setup to provide control authority. iNav is used as a flight controller, chosen due to its versatile motor mixing settings. The craft was built to test its ability at recovery from freefall, as a follow-on from earlier attempts at building a brushless “rocket” craft.

Performance is surprisingly good for what is fundamentally two props on a stick. Initial tests didn’t quite manage a successful recovery, but the repaired single-gimbal version almost achieves the feat. Multirotors in general struggle with freefall recovery, so more research in this area is definitely worthwhile. 

Read more…
3D Robotics

3689739488?profile=originalFeatures and benefits
• Modular parts
o Easy-to-replace powertrain. No more soldering hassles.
• Fully assembled and ready-to-fly
o Avoid the headache and uncertainty of building a drone from scratch. Draco-R is fully assembled, tested, and ready to fly, right out of the box.
• Integrated Jetson TX2 (or Nano)
o Auvidea J120 carrier board
o Draco-R’s flight controller is tightly integrated with the Jetson onboard computer.
• Powerful powertrain
o Inherits the speed of the award-winning Draco racing drone, Draco-R easily reaches speeds of over 100km/h
• Smart battery
o The SMBUS-based battery provides you with detailed information without calibrating the current sensor, unlike with most drone batteries
• Payload mount bay
o Draco-R gives you a mounting plate for a wide variety of payloads
• RTK-GNSS ready
o Draco-R can be equipped with RTK-GNSS (optional)
• Computer vision ready
o Draco-R is tested with most Intel RealSense products including VIO-based indoor navigation and obstacle avoidance.
• Hot swap between desktop and field
o Keep implementing your ideas without interruption! Draco-R supports hot swap and online battery charging while you run a companion computer on your desktop

More information and purchasing details here

Read more…
3D Robotics

DIY Drones now at 90,000 members!

3689739160?profile=original

It's customary and traditional that we celebrate the addition of every 1,000 new members here and share the traffic stats. We've now passed 90,000 members!

A few observations on this milestone:

  • It took 12 years to get to 90k members. It's a marathon, not a sprint
  • Although this is a zero-person operation (it runs itself, with a few volunteers moderating), it's one of the largest robotics communities in the world.
  • Growth has slowed as the drones world has become more Off The Shelf and less Do It Yourself, but we still get a quarter-million pageviews a month. That's not bad!
  • One of the good things about having been so prominent in drones for more than a decade is that our search-engine authority is great. That's why 80% of our visits are from new users -- they're coming via search. 

Once we hit 100k I'm probably going to have to find a better way to celebrate ;-)

Read more…
3D Robotics

3689738601?profile=original

As an industry, we had high hopes for the use of drones in anti-poaching efforts. But in hot climates everything looks hot to airborne thermal cameras and fleet logistics are crippling. Dogs work much better. Here's a sobering report from the field:

I spent 2 years flying anti-poaching drones in the Kruger National Park and other reserves, clocking up hundreds of hours and Robert is right, everything is a trade-off unfortunately. And that’s only the start, there are so many difficulties in the actual operation, very few people have even the slightest idea of what it entails.

The design and building is really the easy part. So if that is hard then one is going to have a hell of a struggle on-site in the bush in 39°C heat (and that’s at 11pm at night!) not counting all the other challenges.

From my experience I believe, that even with the best intentions and best technology, drones will never have more than a negligible effect on poaching. To put it in context, the Canine teams in the KNP’s best haul was 18 poachers apprehended in one week. Our drone team saw less than 8 poachers in TWO years of night flying, with not one apprehension. So if the drone is not CATCHING an average of 2-5 poachers a WEEK then it’s of no use. And note that poachers get around deterrents easily.

Read more…
3D Robotics

QGroundControl 3.5 released

3689738167?profile=original

Congrats to the QGroundControl team who have just released version 3.5 of my fave GCS. Along with a lot of bug fixes, these are the new features:

  • Overall
    • Added Airmap integration to QGC. OSX build only.
    • Bumped settings version (now 8). This will cause all settings to be reset to defaults.
    • Added Chinese and Turkish localization and partial German localization.
    • Added support for the Taisync 2.4GHz ViUlinx digital HD wireless link.
    • Fix loading of parameters from multiple components. This especially affected WiFi connections.
    • ArduPilot Support for ChibiOS firmware connect and flash.
  • Settings
    • RTK Add support for specifying fixed RTK based station location in Settings/General.
    • GCS Location
      • Added UDP Port option for NMEA GPS Device.
      • GCS heading shown if available
  • Plan
    • Polygons Support loading polygons from SHP files.
    • Fixed Wing Landing Pattern Add stop photo/video support. Defaults to on such that doing an RTL will stop camera.
    • Edit Position dialog Available on polygon vertices.
  • Fly
    • Camera Page Updated support for new MAVLInk camera messages. Camera select, camera mode, start/stop photo/video, storage mangement...
    • Orbit Support for changing rotation direction.
    • Instrument Panel
      • Added ESTIMATOR_STATUS values to new estimatorStatus Vehicle FactGroup. These are now available to display in instrument panel.
      • Make Distance to GCS available for display from instrument panel.
      • Make Heading to Home available for display from instrument panel.
Read more…
3D Robotics

Is it a rocket? A ducted fan-copter? Both?

I dunno which it is, but it's really DIY and interesting, and that's all that counts. From Hackaday:

Quadcopters are familiar, and remote control planes are old hat at this point. However, compact lightweight power systems and electronic flight controllers continue to make new flying vehicles possible. In that vein, [rctestflight] has been experimenting with a brushless electric rocket craft, with interesting results.

The build uses a single large brushless motor in the tail for primary thrust. Four movable vanes provide thrust vectoring capability. To supplement this control a quadcopter was gutted, and its motors rearranged in the nose of the craft to create a secondary set of thrusters which aid stabilization and maneuverability.

The aim is to experiment with a flight regime consisting of vertical takeoff followed by coasting horizontally before returning to a vertical orientation for landing. Preliminary results have been positive, though it was noted that the body of the aircraft is significantly reducing the available thrust from the motors.

It’s a creative design which recalls the SpaceX vertical landing rockets of recent times. We’re excited to see where this project leads, and as we’ve seen before – brushless power can make just about anything fly. Even chocolate.Video after the break.

Read more…
3D Robotics

From IEEE Spektrum:

The NIMBUS Lab at the University of Nebraska has been developing drones that have the unique ability to dig holes in the ground and then fill those holes with sensors. If this sounds like a complicated task, that’s because it is: The drone needs to be able to carry a portable digging system a useful distance, locate a diggable spot, land, verify that the spot it thought was diggable is in fact diggable, dig a hole and install the sensor, and then fly off again.

At IROS late last year, folks from the NIMBUS Lab presented a paper detailing a rather burly quadcopter that could carry an auger with an embedded sensor and use it to place the sensor in the ground (you can see a video of this in action here). And at ISER a few weeks later, they presented another paper on how the drone can autonomously figure out whether it’s digging in a good spot or not.

One of the biggest challenges to a system like this is that by the time you pack in the drilling rig and all the sensors and computers that the drone needs to operate autonomously, you’ll be lucky if the thing will manage to keep itself aloft for more than just a few minutes. This is not particularly useful, since the whole point is to send the drone out to place sensors in areas that you can’t easily get to yourself. What’s needed is a way of extending the drone’s range, and the NIMBUS Lab came up with one: A helicarrier, a parachute, and one of the most bizarrely effective drone deployment systems I’ve ever seen.

Read more…
3D Robotics

Ion drive drone -- NO moving parts!

3689736514?profile=original

From Ars Technica. I played with ion drives a long time ago, but never got anything this large to lift. Great work!

The Johnson Indoor Track at MIT probably won't go down in history in the same way as Kitty Hawk has, but it was the scene of a first in powered flight. A team of researchers has managed to build the first aircraft powered by an ionic wind, a propulsion system that requires no moving parts. While the flight took place using a small drone, the researchers' calculations suggest that the efficiency of the design would double simply by building a larger craft.

Ionic wind

In conventional aircraft, air is pushed around by moving parts, either propellers or the turbines within jet engines. But we've known for a while that it's also possible to use electrical fields to push air around.

The challenge is that air is largely made of uncharged molecules that don't respond to electric fields. But at sufficiently high voltages, it's possible to ionize the nitrogen and oxygen that make up our atmosphere, just as lightning does all the time. The electrons that are liberated speed away, collide with other molecules, and ionize some of them as well. If this takes place in an electric field, all those ions will start moving to the appropriate electrode. In the process, they'll collide with neutral molecules and push them along. The resulting bulk movement of atmospheric molecules is called an ionic wind.

Calculations done decades ago, however, suggested that it wasn't possible to generate a practical amount of thrust using an ionic wind. Given advances in batteries, electronics, and materials, however, a team from MIT decided the time may have come to revisit the issue.

Doing so requires navigating a large series of trade-offs. For example, the lower the electric field strength of an ionic wind drive, the more thrust you get for a given power. Of course, if you drop the field strength enough, nothing will get ionized in the first place. Since the thrust per unit area is small, a more extensive thruster system makes sense—other than the fact that it will add to the drag and slow the craft down.

Still, after playing around with different thruster designs, the researchers found that it should be possible to generate sufficient thrust to get something airborne: "This level of performance suggested that steady-level flight of a fixed-wing unmanned aircraft might be feasible but at the limit of what is technologically possible using current materials and power electronics technology."

Finding a balance

The design they chose has a thin wire as its leading edge, where nitrogen and oxygen get ionized. Trailing behind that is a thin airfoil covered by the second electrode. This can both provide a little additional lift and allow the generation of an electric field that accelerates the ionized molecules from the wire to the foil.

But this design had to be integrated with the battery and electronics that make it work, as well as the wing and body that turned the whole thing into an aircraft. Some of those ingredients weren't even available until the team set to designing them.

"Weight constraints necessitated the design and construction of both a custom battery stack and a custom high-voltage power converter," the researchers write, "which stepped up the battery voltage to 40 kilovolts." To handle the aircraft's body, they fed a computer algorithm with a list of their constraints and had it optimize these to allow for stable flight with a limit on the potential wingspan.

The resulting hardware included a five-meter wing with a thin body containing the battery and electronics suspended below it before trailing off to a tail. On either side of the body, hanging off the wing, was a series of the wire/airfoil ionizers (two rows from front to back, both in a column of four for a total of eight). The whole thing weighed just under 2.5kg.

Read more…
3D Robotics

Ten years ago, when I started this site, we were solving some hard technical problems in aerial robotics (such as getting drones to actually fly without crashing!). Now those problems are largely solved (and many of us went on to found companies that today use drone data, rather than making drones themselves), my inner geek took me to the next set of hard technical problems, which are mostly in autonomous cars, AI and computer vision. 

So a few years ago I founded our sister community, DIY Robocars, and today it's more than 10,000 participants around the world doing race and hackathons nearly every weekend. Above is an example (from yesterday's race in Oakland) of the sort of performance we're now seeing -- scale speeds of more than 100 MPH, with advanced computer vision, localization and navigation running at 60FPS on cars that cost less than $200. 

Read more…
3D Robotics

DIY Drones now at 89,000 members!

3689734904?profile=original

It's customary and traditional that we celebrate the addition of every 1,000 new members here and share the traffic stats. We've now passed 89,000 members!

Thanks as always to all the community members who make this growth possible, and especially to the administrators and moderators who approve new members, blog posts and otherwise respond to questions and keep the website running smoothly.

Read more…
3D Robotics

3689734952?profile=original

From Hackaday:

Most civilized nations ban the use of landmines because they kill indiscriminately, and for years after they are planted. However, they are still used in many places around the world, and people are still left trying to find better ways to find and remove them. This group is looking at an interesting new approach: using ground-penetrating radar from a drone [PDF link]. The idea is that you send out a radio signal, which penetrates into the ground and bounces off any objects in there. By analyzing the reflected signal, so the theory goes, you can see objects underground. Of course, it gets a bit more complicated than that (especially when signals get reflected by the surface and other objects), but it’s a well-established technique even though this is the first time we’ve seen it mounted on a drone. It’s a great idea: the drone allows you to have the transmitting and receiving antennas separated with both mounted on pole extensions, meaning that the radio platform can move. Combined with a pre-planned flight, and we’re looking at a system that can fly over an area, scan what is under the ground, and store the data for analysis.

This team includes [J. Rodriguez], [C. Castiblanco] [I. Mondragon] and  [J. Colorado] at the Pontificia Universidad Javeriana in Bogotoa, Colombia. This team attached an Ettus URP B210 SDR card with two Vivaldi antennas to an Astec Firefly drone, linked via WiFi to a Linux server for the heavy data analysis in GNURadio. The two antennas were located on either side of the drone,  attached to a crossbar that separated them and also held the Ettus SDR device.

The trials on the device look promising: the team was able to detect several metal objects in a number of different soil types. The soil type and moisture level is a big part of the problem here: it affects the transmission of the radio signal, and thus the object detection. It’s a promising project that has potential: as the writers note, the use of an SDR system means that the radio detection system can be reconfigured, literally, on the fly. That means it could be adapted to different soil types, and even to detect different sized objects.

Read more…