Blogs

2017 was our tenth year, and since one drone year is like ten regular years, we're practically a century old! A few thoughts on the past twelve months:

First, we've continued to grow here, although not as fast as the boom years from 2012 to 2015 (see above)

More importantly, the two open source software projects that spun out of this community, ArduPilot/APM and Dronecode/PX4, are both going from strength to strength and are now essentially unrivaled in the full-stack open drone software market. APM is thriving with DIY'er (thanks to its developer-friendly "copyleft" GPL V3 licence) while PX4 is thriving with companies (thanks to its corporate-friendly "permissive" BSD licence). 

Both are very active and have large developer bases. Below are the stats for the past month, where PX4 was slightly ahead, but in another month it could go the other way. Although I co-founded APM, I now only contribute to the Dronecode/PX4 project (professionally I can't use GPL code). But I'm super proud of them both, and when it comes to autonomous cars, where I'm still just a hobbyist, I use and love the ArduRover code when we're doing outdoors races that require GPS.

Elsewhere in the industry, the consolidation in the consumer space continued, with eHang the latest to leave and most others cutting prices dramatically as DJI's Mavic proves a hit.  (3DR was one of the first to leave the consumer market, last making Solo in 2015, which is a decision I'm happier about every day, as painful as it was at the time). I predict that prices and margins will continue to fall in that market, and DJI will continue to pull ahead with its more than 11,000 employees. 

The commercial side of the market continued to mature, with most companies now having defined their industry verticals and partnerships and focusing on data quality and the analytical tools expected in pro-level enterprise software.  Most founding CEOs of the bigger commercial drone companies have been replaced by enterprise sales vets, including Airware, Precision Hawk, Kespry, CyPhy, and Measure. (Exceptions are Skycatch, Drone Deploy and my own 3DR -- we must be doing something right!)

It's fair to say that the DIY-centric era of the industry is now over, thanks to today's polished consumer and commercial offerings (we can take a lot of credit for birthing those!). A lot of the DIY energy has shifted to "drone" racing (not actually drones, since they're manually piloted), which does not require the sophisticated autonomy and full-stack software we focus on here. That said, there are still lots of students DIYing to learn how drones work and it continues to be a fun hobby so we'll be here for many years to come. 

Personally, my professional life is totally focused on using drones in the AEC (construction) and GIS (geospatial) markets with 3DR, which had a record year as a software company in 2017. And now that getting drones to fly well is a pretty much solved problem, my hobby/DIY interests have moved to autonomous cars, where the technical challenges are much harder, from computer vision to machine learning. So I spend more time in our sister community, DIY Robocars, than here.  Ten years is a long time to sprint, and the early years here did indeed feel all-consuming. Now we're settled back to a more sustainable pace here, and I do my crazy hacking on things that have wheels, not wings and props.  Plus ça change.

Eagle Eye-10IE is a 3-axis high stabilized gimbal with EO(electro-optical) and IR(infrared) sensors for drone inspection,surveillance,search and rescue applications.The 10X zoom camera provides 1080P 60FPS full HD video streaming and up to 120X zoom capability,which will enable you to capture every detail in the air.While the thermal camera allows you to detect targets in the night or low visibile conditions,and provides you with priceless information when you are carrying out missions like inspecting or monitoring objects.High-performance 3-axis gimbal is developed based on FOC(field-oriented control) technology,which will give you crystal clear video footage.

Eagle Eye-10IE is using the most advanced tracking technology,and it works on both daylight sensor and thermal sensor,with which you can track targets freely during daytime and night time.How it works: put the object you want to track in the screen,move the cursor to the object and lock it via controller,and then the camera will track it automatically.

Another big feature for Eagle Eye-10IE is that the gimbal will geotag the video streaming and photos,also gimbal Yaw/Pitch angle,height,time information will be displayed on camera OSD,this is really important when you are conducting missions like rescue and search.

Eagle Eye-10IE provides an outstanding wiring hub design for RC receiver and video output port,which makes wiring pretty easily.Also the gimbal offers 2 smart speed modes:FAST speed and LOW speed.Fast speed mode is used for small zooming range,which makes the gimbal control sensitive and quick.LOW speed mode is used for large zoom range,will enable you to target the object more accurately.

Feature:

1) Weight:879grams(cameras included)
2) 1/2.9 inch 2MP CMOS sensor
3) 10X optical zoom,12X digital zoom,1080P/60 HDMI output for video downlink
4) 1080P/30 H.264 video recorded for on-board TF card
5) Thermal camera:640*480,25HZ,19mm lens,2X/4X digital zooming,radiometric
6) Auto object tracking function
7) Geotagging,gimbal Yaw/Pitch angle info,height,time will be displayed
7) PWM control and serial command control
8) Convenient wiring hub for RC receiver and video output
9) 3-aixs high stabilized gimbal system
10) Adjustable control speed: SLOW speed for large zoom range,accurate.FAST speed for small zoom range,sensitive and quick.
11) One key back to home position

Product link

http://www.yangdaonline.com/eagle-eye-10ie-10x-eo-ir-dual-sensor-drone-zoom-camera-with-tracking-and-geotagging/

Product photos

Testing Virtual Reality Joystick example based on IMU 10-DOF high speed sytem with APM 2.x board.
NO HOST DRIVERS needed, this work directly on any platform!

This example will be available on the URUS github repo, right now we have pushed 70% code. We hope to push it on the next week.

We have opened a gitter channel for developers https://gitter.im/UrusTeam/URUS and a FB Group for updates and news at https://www.facebook.com/groups/1804355869803790/

Hi Guys,

In, this video, I test and review the most powerful LED flashlight that exist on the market right now : the IMALENT DX80,

also known as "The End of Darkness", producing an insane 32.000 Lumens !

This flashlight contains its own battery pack (4S2P). Its reflector is designed to be both a good thrower (more than 700 meters) and a good flooder (more than 125 degrees) which makes it a good candidate as the most powerful search and rescue drone light.

Of course you will not be able to use this on your DJI play toy as you need to carry this 1.2Kg payload with enough endurance for a SAR mission.

So build yourself a serious X8 multicopter with Arducopter!

Cheers,

Hugues

Wanna see more videos of my YouTube videos ? : https://www.youtube.com/c/HuguesD       

From the New York Times: 

Hi, guys,

A few of our engineers from UgCS team recently built a project in their off-time that we thought some of you might be interested in seeing. 

Since kids having lots of Lego bricks can cause somewhat of a chaos, the guys from our team have built an automated lego sorter to deal with this problem. It uses a vibratory bowl feeder and an artificial neural network (ANN) to feed the bricks up a path, identify them and sort. Currently, in the prototype version, it's only working with maximum brick sizes of 4x2, but with some modifications it can be made to accept a wider range of Lego bricks.

Here's the video of it in action:

We're curious to hear your thoughts on it and maybe some ideas on where it could be used or how it can be improved.

If you are interested in development of this prototype to a more advanced device, feel free to drop us an e-mail at ugcs@ugcs.com

Get the newest version of UgCS here: https://www.ugcs.com/

Wishing you a happy Holiday Season and drone-ful 2018, 
UgCS Team

"Donny the Drone" is voiced by Guy Pearce in this new sci-fi short featuring a modified DJI Phantom.

From IO9:

The film opens in 2022, as Donny—a mapping drone who “woke up” after a midair collision with a bird, and now speaks with the lulling cadence of a new age guru—is being presented with a “Person of the Year” award. Since his transformation, he explains, he’s devoted himself to being an ambassador for machines’ ability to help people. But despite Donny’s big award, not everybody embraces his philosophy—or his vision for humanity’s future.

We’ve seen “the trouble with sentient machines” done many times before, but Donny the Drone offers a thoughtful new take on that well-trod scifi territory. Pearce, recently seen playing android and AI innovator Peter Weyland in Prometheus and Alien: Covenant, is a particularly apt choice to voice the drone who gains a soul—and a slippery agenda along with it.

Hello everyone! It’s been a while since I write a blog post here!

In this post I will explain about a very cool robotics project I did last year. This project is as cool as it sounds: making a drone follow a roomba using computer vision!

This has been done before and for sure someone has posted their on solution here, I will just describe my version.

The idea can be broken down in different sections, one the vehicle, the communications, the control, computer vision and results…

The objective of this project is to make a drone “see” a roomba and try to follow it from the air.

On the vehicle side, I was using the great and cool BigX, which is a great vehicle with very nice performance and big!! Here is a pic of it:

On board I was using the same concept of the Flight Stack, which is the combination of a flight controller (pixhack in this case) and a raspberry pi with extra features.

The RPI was a 3, and its the one in charge of “piloting” the aircraft when I’m not doing it. Also the RPI is the one that runs the computer vision algorithm to be able to “see” the roomba. With that information (position in pixels X and Y) the RPI will compute the necessary velocity commands to steer the vehicle to the center of the target.

The RPI was also in charge of the communications, it created a network and on the ground I was using a ubiquiti AP that was connected via ethernet to my MBPR, I used this configuration because such AP gave me the range of 10km LoS (I just try it up to 3km…)

Also on board, connected to the RPI, a B101 HDMI bridge was used in order to capture the frames from the gopro camera and have them analyzed.

On the ground side as I mention before, I had the Ubiquiti AP and my laptop connected with ethernet to it. My computer logged in to the RPI via SSH in order to activate the scripts that run the main stuff. Also I had QgroundControl opened to be able to see the telemetry of the vehicle in a nice way. I was using mavproxy with udp casts. An image of how my computer screen looked was:

In the image above you can see the position teleoperation tool from the AltaX ground station program. This module changed the position of the robot by reading the keyboard from the ground station computer, pretty neat…

On the computer vision part, I added blue tape to the top of the roomba in order to be very easy distinguishable from the environment. I also tuned as much as possible my different color tracker algorithms, you can find the code here and a video demonstrating here.

When you combine all ingredients, plus creating a velocity vector position control, you got a nice result, like the one showed in the video showed at the beginning. 

Let me know what you think, in this case I cannot release the as its protected by my employer but as usual I can guide or make recommendations to anyone doing something similar.

Don't forget to fly safe!

Hi all, 

We have been trying to achieve this goal for a long time. Two years ago we do a flight of 3 hours and 10 minutes with very early model, and our next goal was 4 hours but the guys of Skyfront published a record of 4 hours and 34 minutes. Now we are super happy to announce that we have surpassed this record!!!

We think it is the longest flight ever done with a multicopter (not including thethered drones!)

For those who do not know the project, it is a hybrid quadcopter, it is an electric multicopter that has an on-board generator based on a small 2-stroke engine. The unit that has been used to this flight is a standard HYBRiX in which the original fuel tank has been replaced by a larger one.

I remember that at the beginning there were many doubts about the technical feasibility of this configuration.Today this is the look of our product:

I wanted to thank everybody who was involved in this project.

Best regards.

jlcortex

    Our customer from Malaysia enjoys some photos of Skywalker X5PRO with whole set to us and I am really pleased to share with you. I think it is really terrific airplane to do professional mapping survey and outdoor task considering its intelligent operation and high efficiency. 

After preparing work, they start to take off this aircraft. It is really intelligent and fly faster.

The link of Skywalker X5PRO with whole set:

https://www.youtube.com/watch?v=zJLW-7HzB84

I have written blog of Skywalker X5PRO with whole set before and you could see more information in the video. Also, in terms of interface of ground station, you could see the picture that we have English version.

https://www.youtube.com/watch?v=czcJ2tAXUr4

If you have any question, you could contact with me. My skype is skywalker-ellen@outlook.com. My email is skywalker-ellen@outlook.com.

Hey everyone,

We are pleased to announce Emlid Edge drone controllers are in stock! The time has come to get your drone and your action camera prepared for flight!

Emlid Edge is the first autopilot platform to feature an integrated HDMI input and video compression engine allowing it to capture Full HD video from an onboard camera such as GoPro or any other HDMI-capable camera. Video is streamed directly to the Ground Control Station over a long-range 5.8 GHz Wi-Fi link.

Also, check out our Youtube channel where we continue sharing Edge demo flights! 

Learn more about Edge: https://emlid.com/edge

     Recently, Skywalker launch New Skywalker X8 that is made of new composite material instead of EPO material, aiming at satisfying customer's needs for professional mapping survey.

You could see the picture of this special material of Skywalker X8. No matter on the aspect of manufacturing technique or intensity, it has total different from former Skywalker X8. To a greater degree, it has improve a lot on flight performance.
I will introduce the specifications of this Skywalker X8(honey comb composite material).

If you have any question, you could contact with me at skywalker-ellen@outlook.com.

Hi everybody, this is my last short tutorial for setting up a Raspberry Pi as Companion PC.

That is easy stuff for experts, but definitely something I would have appreciated when I started.

Crazy good deal on the best autopilot for small drones and rovers. This deal won't last long.  Free shipping for Prime members.

Includes GPS, power module, 8-channel servo rail expander for planes and designed for latest Dronecode/PX4 and APM software. 

From Hackaday:

The commission just levied a $180,000 fine on a company in Florida for selling audio/visual transmitters that use the ham bands as well as other frequencies.

The FCC charged that Lumenier Holdco LLC (formerly known as FPV Manuals LLC) was marketing uncertified transmitters some of which exceeded the 1-W power limit for ham transmitters used on model craft.

Equipment that is purely for ham use is normally exempt from certification, but since the equipment was able to operate on other frequencies, this was a violation. In addition, even for licensed ham use, some of the transmitters were using too much power.

The company stopped selling the units in question after an FCC inquiry back in April. We can’t help but think that in years past building a consumer product with a significant radio transmitter was a big task, and someone would bring up the FCC rules and certifications before much progress had been made. These days though you can easily acquire building block ICs and modules to field a product in a few weeks that would have taken a sophisticated team years of effort not long ago.

Hi, guys,

We'd like to update you on what's happening with our universal adapter for DJI cameras (see here for more info in case you've missed it).

While developing our tethered drone system AirMast, we've also put work into designing a universal camera adapter for DJI cameras. What this means is that using it you are able to use DJI cameras and gimbals on almost any drone or even without the drone itself.

Currently the following camera gimbals are supported: DJI Zenmuse: X3, X5, X5R, XT, Z3

We do not plan on selling the hardware so we've made that open source and have created a GitHub documentation for assembling it: http://doc.airma.st/camadapter/

For controlling the camera you need software for the onboard computer and the web-app for controlling the camera. We've made it available for purchase from our site here: https://www.ugcs.com/en/page/airmast-camadapter

If you want to get in touch with us or have some questions regarding the camera adapter, either leave a comment below or drop us an e-mail at ugcs@ugcs.com

Wishing you a warm and happy Holiday Season,
UgCS Team

Last year at this time, I reflected back on the news and trends of the commercial drone markets of 2016 and wrote about the mixed state of affairs ahead for 2017. Throughout the year, I offered my perspective on how the drone industry was still motivated by hype and how assessing forward momentum required hard data on the performance of the various sectors of the industry. To that end, we did research over the summer that surveyed 2,600 respondents on drone purchases, service providers, business users, and software services. In September, we published the data in 2017 Drone Market Sector Report 2017.

In this post, I’ll use that data to illustrate the major trends of the past year and describe what I think are the major challenges ahead for the drone industry.

Listen to this companion Drone Radio Show podcast here for the complete assessment.

Trend 1—Growth

By all measures, the drone industry in 2017 was marked by significant growth – growth in aircraft sales, software licenses, the number of service businesses entering the market, and the number of industrial businesses setting up commercial operations.

Here are a few statistics:

• We project U.S. sales in 2017 to be about 3.3M units, which is 36% above 2016 figures. That’s all drones, all sizes. It’s about 1.3M units for the >250gram category.
• As of October 31st, there were about 837,000 hobbyist users and 107,000 non-hobbyist drones registered with the Federal Aviation Administration (FAA).
• As of December 1^st, there were about 66,000 Part 107 FAA Pilots.

This represents a big change in the commercial market since Part 107 regulations supplanted Section 333 as the means for commercial operations in the U.S. What this and our survey data tells us is the number of service providers currently outpaces demand, and as a result, service prices are coming down significantly.

Trend 2—Consumerization

We said in our report that more consumer drones are being used for commercial work than ever before. For example, our data shows that more than two-thirds (68%) of all drones weighing over 250 grams are purchased for professional purposes—either governmental or business.

Why is this significant? Because the impact of consumer-originated technology on the enterprise is something that can’t be ignored. Enterprises want to take advantage of powerful, yet easy-to-use products (like DJI’s popular consumer models), and put them to work on the job. What this means for operators or businesses is that a shared core technology benefits all users and enables companies to scale the best experiences to everyone. Enterprise customers get the added simplicity and usability of the consumer product that has been built to meet the demands of thousands of customers around the world.  The average individual pilot gets to benefit from the reliability and scalability inherent in the product and demanded by enterprise users.

Trend 3—The DJI effect

Our data shows DJI is the clear market leader in drone aircraft sales and almost every software category. For example, DJI is the dominant brand for drone aircraft purchases, with a 72% global market share across all price points and an even higher share (87%) of the core $1,000–$1,999 price segment. Additionally, in the three categories of software we evaluated, DJI is the market-share leader in two: flight logging and operations, and automated mission planning.

This is significant because by building on top of its existing technology platform, DJI has fast-tracked development and has benefited from economies of scale. By migrating a successful technology stack and feature set up market, DJI never has to reinvent the wheel—it just needs to improve upon the original design and save engineering cycles for real innovation.

The upshot is that, to stay relevant, all the other major vendors have had to partner with DJI (see Trend 5 Partnerships, below). DJI’s sales success has taken market share from others and has led to layoffs at 3DR, Autel, GoPro, Parrot, and Yuneec. However, fears about data security remain. And this has some speculating about whether DJI can sustain its leadership role in the future.

Trend 4—Investments

According to CB Insights, investments shifted in 2017 from aircraft hardware to software. In 2016, there were 106 deals totaling $542M. Most of these were for hardware. In 2017, VCs focused on software, end-to-end solutions, and counter-drone technology. CB Insights projects the year will end with 110 deals totaling $494M. The most significant investment this past year was 3D Robotics’ $53M Series D round. It saw them pivot from hardware to software services.

Why is this significant?  Because it shows the industry is still maturing. Seed and Series A rounds represented 60% of all deals in 2017; whereas early-stage share peaked in 2015 at 73% of deals. Additionally some of the most well-funded drone companies are targeting enterprise and industrial inspection.

What this means for operators or businesses is greater affordability. Software advances, computer chip manufacturing techniques, and economies of scale will continue to drive down the cost of drone platforms and sensors and solutions.

Trend 5—Partnerships

This year we saw a change from synergistic merger and acquisitions to the creation of end-to-end solutions via partnerships. For example, look at how DJI’s enterprise partnerships have grown. Consider their AirWorks conference. What drone major vendor wasn’t there? The list included DroneDeploy, Measure, PrecisionHawk, Skycatch, and Sentera, to name a few.

This past year we also saw an uptick in regulators and industry stakeholder partnerships. For example, the Drone Advisory Committee was created to provide the FAA with advice on unmanned aircraft integration from a diverse group of stakeholders. Major commercial participants include Intel, DJI, Amazon, Google X, and Facebook, as well the Aircraft Owners and Pilots Association.

Consider also the FAA’s new UAS Integration Pilot Program. Here, government entities are partnering with private-sector companies, such as unmanned aircraft systems (UAS) operators and manufacturers, to submit proposals to the FAA to fly more advanced operations in their communities, including flying beyond line of sight and over people. This is significant because it’s clear that regulators want to include industry when creating policies.

However there is some good news / bad news with this.

The good news is greater flexibility. With vendor partnerships, drones will be able to perform more types of data gathering in a shorter timeframes and with more precision than many other options. So, more aircraft, sensor, and software integration.

The bad news is operators and businesses have regulatory uncertainty. We advise our clients to plan for some uncertainty as technology, the public, and bureaucracy find common ground on operations for beyond visual line of sight and over people. 

Challenges ahead

Here’s my list of the major challenges facing the drone industry in 2018:

1. Regulations: We may see more regulatory red tape—e.g., a patchwork quilt of rules as the FAA’s UAS Integration Pilot Program begins to make policy.
2. Public sentiment: Basic public concerns still exist about drone safety, security, privacy, and their public nuisance. My question is: How can we overcome this?
3. Business value: We’ve yet to see credible ROI that hits the executive scorecard. The key question is: What monetary benefit do drones and information gleaned from drones provide shareholder value?
4. Information accuracy: Up to now, drone vendors have been focused on the accuracy of image capture and the rigor of the drone system. For better business value, they need to focus on the accuracy of the data processing and the rigor of data analysis.
5. IT data governance: This is especially the case for drone inspections where a single drone could collect 50 to 100 gigabytes of data. Managing these large data sets starts to become one of the things that has to be worked out.
6. Automation: A lot of software automation will come, including artificial intelligence (AI) or algorithms that minimize the amount of human effort to distill all that information and get to some actionable inference. But large scale industrial use of AI is young and it requires manual intervention to distinguish the difference between near-similar objects.
7. Endurance: We’re still on the quest for efficiencies like better power sources or mixes.
8. Widespread business adoption: Business and industry adoption is growing, but it’s mixed because of factors such as business risk aversion, concerns over invasion of privacy, and a reluctance by many companies to share too much information about successes.

That’s it for now.

Listen to the companion podcast here http://bit.ly/2CXe6uK.

Look for a follow-up piece on our specific predictions for 2018, which will include investments, technology improvements, ecosystem partnerships, and software innovations.

If you have questions about what's in the report I mention or would like to comment, write me at colin@droneanalyst.com.

Image credit: Shutterstock

Last week I received from Gens Tattu two Li-Po batteries for review. The Tattu 2s 800 mAh and Tattu 3s 450 mAh. Firstly, I thought about making an individually test of each battery, but latter I found it more helpful to make a side by side comparison of them. Both Li-Po batteries are suitable for micro class FPV drones like the Blade Torrent 110 and KingKong ET series. Weighing about the same (~42 grams), I’m curious to find out which one will perform better (top speed vs flight time).

For my tests I will use the KingKong ET125 drone, which weights about 120 grams and it is equipped with 4-in-1, 10 Amps ESC (2-3s rated) and 7500 KV brush-less motors. Roughly, the number before “KV” represents RPM/Volt (under no load!), so higher battery voltage means more RPM, respectively higher speed.

In order to obtain higher voltage, Li-Po packs are made from several 3.7 Volt cells, serially connected. Therefore, a 2s LiPo (2 cell) has nominal voltage of 7.4 V and a 3s (3 cell) one has nominal voltage of 11.1 V.

Which battery is best for micro racing drones, 2s or 3s one?

Both batteries came well packaged and, as I requested, with JST discharge plug. While both LIPOs weights about 42 grams, they have a bit different form factor. As usual for Tattu batteries, they came half charged, about 3.82 V/cell.

Tattu 800mAh 2S lipo battery:

Specs

• Number of cells: 2 (2s);
• Voltage: 7.4 V (3.7/cell);
• Capacity: 800 mAh;
• Max Continuous Discharge: 45C (36A);
• Discharge Plug (drone side): JST-SYP-2P;
• Max Burst Discharge: 90C (72A);
• Dimensions: 51.6 x 25.6 x 21 mm;
• Weight: 42 grams.

As the stock battery of the KingKong ET125 has only 550mAh, I’m expecting at least 1 minute bonus of flight time for the extra 250mAh. Weight difference between 550mAh and 800mAh is about 7 grams only.

Specs

• Number of cells: 3 (3s);
• Voltage: 11.1 V (3.7/cell);
• Capacity: 450 mAh;
• Max Continuous Discharge: 45C (20.25A);
• Discharge Plug (drone side): JST-SYP-2P;
• Max Burst Discharge: 90C (40.5A);
• Dimensions: 54 x 29 x 13 mm;
• Weight: 41 grams.

Tattu 2s 800 mAh vs Tattu 3s 450 mAh

First off all, powering your drone with higher voltage batteries (like 3s instead of 2s) can produce permanent failure of electric parts. My suggestion is to check the “s” ratting of each individual component. For example, the motors and ESCs could be rated for 3s, but the VTX only for 2s.

After five charging/discharging cycles, I fully charged the batteries at 1 Amp. As you can see in the comparison table bellow, both LIPOs ended well balanced.

Real world test

Next day, with confidence I went out to do some real world test. Firstly, I loaded the 2s LiPo into the KingKong ET125. As you can see in the video bellow, due the strong wind gusts, my flight was pretty erratic, so the results are not too relevant. I need to repeat it in more calm day.

Although, KingKong clearly says that is risky to use 3S LiPo batteries with the ET125, it worked very well. I noticed some improvements, but you need to be a damn good pilot to take advantage of the extra speed provided by the Tattu 3S 450 mAh. IMHO, 7500 KV motors are just fine with 2s batteries, especially with a light weight frame.

Tattu’s “Less is more” aka “Less weight, more flight time” slogan, proved to be pretty fair. While their 2s 800mAh LiPo has ~19.05 mAh/gram storing capacity, the 550 mAh Gaoneng LiPo has only ~15.71 mAh/gram. Compared to stock battery I got more than 1 minute extra flight time.

Originally developed as an MAV flight simulator, Microsoft's AirSim program recently added the ability to drive a simulated  car.  Although my own interests lie on the MAV side,  students in my recent AI / Deep Learning course chose to work with the car simulator as the basis for their project. 

As the video and code show, AirSim's support for Python interaction, combined with the powerful features of the TensorFlow deep-learning package, made it a matter of a few days' work to put together a simple demonstration of visually-guided collision avoidance. 

With sophisticated simulation and machine-learning tools available to anyone for the price of a decent workstation and GPU, it promises to be an exciting year ahead!

What is MAAXX Europe?

MAAXX challenges competitors to design, build and race autonomous drones against each other. The rules are simple, follow the marked track as quickly as possible with your aircraft fitting in a 1X1X1m cube, and make sure the drone can race by itself. You can either build your own drone, or take part in our Hackathon using a DJI Spark.

The event is on March 23-24th, held at the University of the West of England in Bristol.

How do I enter?

1. Register online at www.maaxx-europe.com
2. Submit your aircraft profile
3. Turn up and race!

What's up for grabs?

There are four prizes up for grabs this year:

• Speed prize - Fastest 10 consecutive laps
• Endurance prize - Most laps completed through the event
• Hackathon Cup - Fastest 10 consecutive laps completed using our Hackathon DJI Spark
• Kudos prize - Coolest team as nominated by the other teams

Want to know more? Comment here, visit www.maaxx-europe.com or email Miles.IstedsJacob@UWE.ac.uk