Hi everyone!I was curious to know if anyone would be interested in participating in a challenge to make something AWESOME with a new breed of processor which is particularly great for DSP/networking/USB/motor control type applications or more basic things if that is overkill for what you want.XMOS event driven processors allow you to execute code in parallel which could open up a number of possibilities to evolve your drones "mind" to the next level!I have some development kits to give away potentially, but before I give away all the details, I just wanted to see if there was any interest before I add to the challenges area or such.People have made some cool stuff with our processors before - check out our official YouTube channel here to view some cool videos. Feel free to subscribe to be kept updated with latest videos and XMOS news.Some details of the processor itself which will be on the dev kit (to be released soon):- Single core device (Although we do have quad core versions - ask me)- 400 MIPS per core.- 8 Threads per core.- 64Kb RAM- 8KBytes OTP memory for applications, boot code or security keys, with security mode- 64 user I/O pins- Support for high performance DSP (32 x 32 → 64bit MAC) and cryptographic functions- Time aware ports provide up to 10ns timing resolution- Designs implemented using a software-based design flow (can program in C or XC - very similar to C but with support for extras such as parallel execution)- Scalable - can connect many kits together for crazy amounts of processing power using Xlinks.For detailed spec see: http://www.xmos.com/products/xs1-l-family/l1lq128If anyone is interested please feel free to post below with your project ideas and answer the 5 questions at the end of this post (and follow/subscribe to this forum topic for updates so I can contact you nearer the release date) so I can see if there is enough interest. I hope I have posted this in the right place, if not please feel free to move this to the right part of the forum!EDIT: As there has been some interest shown (especially in a number of robotics forums) and to save time later - please post below with the following info if you would like to be considered:1) Project Title2) Project Description3) How many Dev Kits you think it may require (eg you may need 2 if you are demonstrating the ability of one robot using image recognition to track another for example)4) Are you willing to keep a video/photo diary of your progress if we choose you? Please state which/both.5) Your skill level/previous experience.
You need to be a member of diydrones to add comments!
Thanks guys for submitting ideas, unfortunately entries closed for this a couple of months ago and kits have been sent out - see http://www.xmos.com/challenge/ . However your propositions are very interesting. I have sent you all a PM with more details.
Jason,
I'm working on building a neural network based autopilot, one that adapts to varying flight conditions, airframes, etc. in real time. A highly parallel processor like this one could be very interesting for what I'm doing. So, here's my submission:
1) ANNpilot
2) An Artificial Neural Network (ANN) based autopilot for small aircraft
3) Anywhere between 1 and 5. Alternatively, a five core kit might work
4) I would love to keep a photo diary, although I don't really have the ability to take quality video.
5) I have built a neural network based autopilot for a 4 rotor helicopter, as well as a robot boat (15 ft long, with radar, LIDAR, and other sensors) and other projects. I'm a senior engineering student, graduating in May.
Title: Autopilot system for high speed FPV flying wing.
Project Description: XMOS processor will be used as the core processor for a custom autopilot system for a high speed flying wing equipped with First Person Video system.
The airframe will be a foam core, carbon fiber spar, fiberglas skin flying wing. It will use a brushless electric power system using high discharge rate lithium polymer batteries.
The First Person Video system will provide a real time video downlink allowing the pilot to remotely pilot the wing and simultaneous video recording.
The autopilot system will provide functionalily for several modes of operation including (1) stabilization - allowing the pilot to fly at high speed and in windy conditions with relative ease, (2) failsafe - providing for a controlled climb to a safe altitude followed by an autonomous return to the take-off area in case of loss of radio signal, (3) Full autonomous mode - allowing the aircraft to fly a prescribed course without pilot interaction. Othe modes such as vehicle tracking, etc. may also be developed.
# of dev kits required: 1 minimum, 2 optimal as there will be two members on the team.
Video/photo diary - Yes, I will keep a combination of video and photos. Me working in my shop for hours on end does not justify video, but test flying certainly does!
Skill level and experience - I have a masters degree in electrical engineering and am involved in the development teams within the DIYDrones community. I have successfully integrated an IMU based autopilot system using pieces of open souce hardware and my own software. My brother (second team member) is a programmer by profession and will help with the software architecture to best use the parallel processing capabilities.
What kind of timeframe should these projects be presented within?
I am working on what could only be described as a complete UAV control system (IMU, motor and radios plus tons of sensors) in one integrated system.
I have been considering several CPU types but sofar I have tended towards ARM7 platforms which I have had a good experience with in the past, but it seems from the information available that this might be an even better solution. Especially since this will need multithreading/concurrent execution to a greater degree than
previous projects as well as cpu speed.
I would be interested in evaluating this but I would also want to know what "strings" would be attached to these dev-kits..
OTP is a big problem in hobby applications. While OTP is OK in a product, without firmware upgrade it's pretty unusable for hobby experiments.
Is it possible to run the code from RAM and using the OTP just for a bootloader from a serial flash?
Replies
I'm working on building a neural network based autopilot, one that adapts to varying flight conditions, airframes, etc. in real time. A highly parallel processor like this one could be very interesting for what I'm doing. So, here's my submission:
1) ANNpilot
2) An Artificial Neural Network (ANN) based autopilot for small aircraft
3) Anywhere between 1 and 5. Alternatively, a five core kit might work
4) I would love to keep a photo diary, although I don't really have the ability to take quality video.
5) I have built a neural network based autopilot for a 4 rotor helicopter, as well as a robot boat (15 ft long, with radar, LIDAR, and other sensors) and other projects. I'm a senior engineering student, graduating in May.
Here is your first entry -
Title: Autopilot system for high speed FPV flying wing.
Project Description: XMOS processor will be used as the core processor for a custom autopilot system for a high speed flying wing equipped with First Person Video system.
The airframe will be a foam core, carbon fiber spar, fiberglas skin flying wing. It will use a brushless electric power system using high discharge rate lithium polymer batteries.
The First Person Video system will provide a real time video downlink allowing the pilot to remotely pilot the wing and simultaneous video recording.
The autopilot system will provide functionalily for several modes of operation including (1) stabilization - allowing the pilot to fly at high speed and in windy conditions with relative ease, (2) failsafe - providing for a controlled climb to a safe altitude followed by an autonomous return to the take-off area in case of loss of radio signal, (3) Full autonomous mode - allowing the aircraft to fly a prescribed course without pilot interaction. Othe modes such as vehicle tracking, etc. may also be developed.
# of dev kits required: 1 minimum, 2 optimal as there will be two members on the team.
Video/photo diary - Yes, I will keep a combination of video and photos. Me working in my shop for hours on end does not justify video, but test flying certainly does!
Skill level and experience - I have a masters degree in electrical engineering and am involved in the development teams within the DIYDrones community. I have successfully integrated an IMU based autopilot system using pieces of open souce hardware and my own software. My brother (second team member) is a programmer by profession and will help with the software architecture to best use the parallel processing capabilities.
PM me with any questions :)
For those of you who are involved in the XMOS Challenge - we have recently launched a new site XCore Exchange - the site to be when developing projects around the.... If you need advice do check it out.
I am working on what could only be described as a complete UAV control system (IMU, motor and radios plus tons of sensors) in one integrated system.
I have been considering several CPU types but sofar I have tended towards ARM7 platforms which I have had a good experience with in the past, but it seems from the information available that this might be an even better solution. Especially since this will need multithreading/concurrent execution to a greater degree than
previous projects as well as cpu speed.
I would be interested in evaluating this but I would also want to know what "strings" would be attached to these dev-kits..
Is it possible to run the code from RAM and using the OTP just for a bootloader from a serial flash?