Hi
My daughter is interested in robotics and is eventually (well less then 2 years away) looking at taking some sort of engineering at university.
After looking at the various different kits, decided that Vex was the one she wanted.
She’s Looking at the autonomous side of things and we have a limited budget.
I have a question, Has the Cortex controller Hardware, either Memory, or Processor Power ever proven restrictive, and is its performance ever an issue.
I have been writing Games since the days of the C64 and Spectrum, and have fitted games including graphics into 32K system with a 4Mhz processor, so know what can be done, although I had to use assembler and machine code at the time. I assume though that anything to do with vision is out.
I also take it that RobotC outputs microcode that is interpreted, does this pose an issue at all.
Looking at the various systems, Cortex Looks on Par with NXT, and VexPro looks on par with the Fischertechnik ARM 9 Micro Controller.
If I ignore the chance to enter competitions, should we be considering Fischertechnik at all, as cost wise I can get a kit with an ARM 9 Micro Controller for less then a standard vex cortex starter kit, or is the Micro Controller less of an issue, and the actual Vex Robot Hardware makes up for this.
The vexpro is a bit out of our price range, however if we managed to get one in the future, can the vexpro controller be used with standard vex hardware?
These are pretty wide open questions, but I can provide some comparative answers. Having been involved with Vex competitions for >4 years 2 years using the PIC and 2 + with the cortex, the cortex is hands down a much preferred platform. I have not seen anyone hit the limits with the code I or my team have written. I have written code to evaluate all the various sensors Vex offers and I have yet to run out of processing power or space even when I’ve had every input / output port utilized.
There were speed tests run comparing RobotC to Easy C on a Cortex a while back and they were very comparable. so the interpretive byte code does not seem to be an issue. I use Robot C as I have found it easier to work with C code and in my opinion the support is better. as a bonus they support many platforms, I have not yet taken advantage of this but they even are supporting the Arduino platform.
NXT looks like a scaled down version of the Cortex. (arm 7 vs 9 ) The main reasons I never picked one up are 1) I was already involved with a Vex team before I got interested in robotics again, and 2) there are just a few sensor / motor ports, it does not look to be too extendable. But it looks great for quick and fun projects.
I started with a PIC based proto bot and added hardware, sensors, and a cortex finally when it came out. I found having the kit hardware and the standardization / compatibility of the Vex system to be a great benefit to productivity in building robots.
I believe Vex Pro is compatible with all Vex H/W but the programming environments are different, I don’t believe there are drivers for all the sensors, so that would need to be considered.
In the end what do you want to get out of the experience? In reviewing the Fishertechnick it is on par with the cortex, but does not seem to have a ready to purchase handset. Also I’m not familiar with the tool chain. Finding support or other users may be more difficult.
Most of my experience have been with working on tele-presence robotics. I get enthused about autonomous, but the competition has not emphasized it as much in the last few years so it is an afterthought for the team. But the students get most pleasure out of driving the robots and controlling the various appendages with the remote control.
Have you looked at the arduino line of products? it offers a fairly powerful platform and is fairly inexpensive. I purchased a Mega 2560 & a couple XBee radio boards to work on a robot to robot communications project this year. I found it very straight forward to get the arduino working with the XBee (I pick the mega as it has multiple H/W serial ports). I have had much more of a challenge getting the Cortex to interface to the XBee Although I think I finally worked it out (with a fair amount of help from the forums here)
I’ll end with a positive endorsement of the Vex Product line. They make very robust equipment, and the company has great support.
Best of luck in your decision, Have a great day. - Kb
It’s unlikely that your daughter would outgrow the cortex in the next couple of years. Memory can be a limitation for some
algorithms but that is likely to be the case with most small embedded processors. Vision is certainly out unless you use an external image processor and send the results via a serial link or other method to the cortex, you could take a look at CMUcam or something similar if necessary.
Not an issue for all practical purposes. It also has the added advantage of the PC emulator and the virtual worlds simulation environment, something that would have been difficult without the intermediate byte code.
I have never used the Fischertechnik controller, I did a quick google and read the specs and feel it falls short of the cortex in some areas. Motor control is weak, only 4 motors with a total draw of 2A spread amongst them. The cortex is designed to handle 10 motors with significantly more power. The cortex is using industry standard (in terms of the hobby RC) connections so any RC servo can be hooked up that falls within the spec for power.
I think you should also consider the companies behind the two products. Fischertechnik is a toy company, VEX is a robotics company targeting students like your daughter and is staffed by people who have been involved with robotics for many years.
I have a VEXpro as well as a cortex, they are different and aimed at different users. The VEXpro is theoretically more powerful but that comes at the expense of a more significant learning curve. If you are a Linux supporter and have other linux boards (Raspberry Pi ?) then perhaps it would be suitable. If your daughter is a beginning programmer, then the cortex is the right product for her.
Although you mentioned that autonomous is the primary interest, I would encourage you to get a bundle that includes the vexnet joystick and WiFi keys. Buying them as part of a bundle does save money compared to adding them later and will allow the ROBOTC debugger to be used when running autonomous code via WiFi (I should add you also need the $50 serial programming cable).
thanks for the answers, I think the thing I need to remember is that its for my daugher and not me.
With the Fischertechnik controller I was blinded by the 8 Meg of Ram, but didn’t notice that it could only handle 4 motors, and the Cortex also seems to have more inputs.
Whats scary about 64k is that if I decided to let the robot map a room onto a 256x256 grid, its gone, of course there are much more efficent ways of doing that, but its just a thought.
Although I have had some linux experience, what with trying to program a little NAS and having to use cygwin as some of the sony and nintendo tools I’ve had to use needed it, again though I must remember its for my daughter, and I don’t want to put her off at the 1st hurdle. It would be nice though, if in a year or so’s time she wanted to do something complex, that needed the VEXPro, that any investment into standard vex hardware wasn’t totally wasted. Also I quite like that the VEXPro seems to have proper wifi.
Thanks for the comment on the controller, although primarily looking at autonomous, being able to debug wirelessly would be a big advantage, also being able to initially control the robot directly would give a quicker reward for her construction efforts, just need to see if my budget can stretch.
Many years ago, I attended (the now defunct) Middlesex Polytechnic, it later became Middlesex University before they scrapped the engineering school. I spent three years in the Wood green and Winchmore Hill area.
Would your plan be to buy the clawbot starter bundle or are you thinking of something else?
Well many years ago I went to Coventry Polytechnic (now university I think)… which is only a bit to the right of Worcestershire
Yes the plan is a clawbot starter bundle, well the 'Programming Control Starter Kit ’ when it is available, and I’m seriously considering adding the joystick and adapter keys to that, which based on the US site, works out cheaper then getting the ‘Dual Control Starter Kit’ and then adding the sensors from the programming kit.
Presumably the Joystick would also offer a way to communicate with the robot when its in autonomous mode
I think that’s a great choice myself. I started with the protobot (smallest thing available) but also got a claw, and we have had a good deal of fun with it. My son lost interest after a bit, but I like to keep a basic clawbot in working order. Add a small wireless camera and you have a fun driving package. You can also add routines that are ‘triggered’ but key presses. It is a very versatile package.
Well the stuff arrived yesterday, and my Daughter was very excited opening it and looking at all the bits and pieces, and got straight into building the claw bot, after first plugging in the battery to charge.
However because of a schools fireworks night, she had got home late so has only managed to do a bit.
The only negative thing was screwing in the motor screws, which I had to do, I also found those difficult and had to apply quite a bit of pressure to keep the hex key in while turning it.
Now I have to wait for her to get home from school today so that I can see this thing working, hopefully she wont have too much homework
I’ve not had much success with the lock screws, I bought a batch last year for the school and have pretty much banned the kids from using them as I find they are hard to install and even harder to remove. They may have improved since then but I have not checked. If there is room I often use a socket head cap screw instead of the standard VEX screws now.
We have good tools. I can only assume that perhaps it was a batch of screws with too much thread locker applied. I know others had problems a few months back. I should try again as I really would prefer that the motors didn’t come loose as they inevitably do with conventional screws. I also wish lock washers were legal despite their reputation for not always working, we use them on non-competition robots and they work well.
Well I’m almost frightened to undo the screws now as I think it may mash the heads.
I think the main issue was that the hex keys didn’t seem to fit the screws properly… if the screw heads were deeper and the hex key could stay in while turning then I’d be happier… maybe I should invest in proper tools.
Personally I think maybe having a Bolt head (not sure if thats the correct term) would have been better, then I could use a standard socket set on them, and maybe you could then apply a cap held in-place by a couple of rivets to stop it turning or something…
anyway my Daughter wrote her 1st code yesterday, just basically converted the tank type control to single stick control and put the claw raise lower on the other stick.
a team can consist of one person … if you are, you are gonna have to register and go to either the birmingham or bristol regional … otherwise you cant qualify for nationals
depending where you live, we are running a small sack attack workshop consisting of 3-5 teams, we are hoping to do it around december the 11th …
it isn’t a competition, it is a workshop where we are going to talk about programming, design and build etc, as well as test the robots etc
if you are interested, email me at [email protected], and i will email you some details after the date is confirmed (Hoping to get it sorted in the next couple of days)
The workshop seems interesting, but I think we’re too far away (I’ve added my location to my profile now) and probably have too much on.
Regarding VRC
We have the London regionals in about 10 days… so no chance for those, especially as thats the day she’s visiting Aero Engine Controls in Birmingham, which is a shame, especially as they only have 3 teams at the moment, also presumably entering an unmodified clawbot is not A good idea.
the Birmingham onces are a couple of months away… so… maybe…but still a bit close for comfort, and should you be from the region to enter the regionals?
I’ve looked at most of the wiki’s but it all seems a bit daunting.
Is there a step by step tutorial/guide/checklist for on competitions somewhere as this is all still very new to me.
Also can you enter sack attack with only 1 robot?
That aside I think we may have to wait until next year
With the VRC, you can only enter one robot per ID, you can register more than one robot … checklists, guides and manuals etc can be found at this URL … https://vexforum.com/wiki/index.php/Sack_Attack (at the bottom)
with tutorials, there are a couple scattered across youtube, but there are the vexroboticstv referee training video which are helpful, i linked them on our site
with competitions, we had two month to prepare for our gateway competition, and we came joint 3rd, with the energy award, i recommend you consider the VRC, it is great fun, if not … look forward to seeing you next year
with your location, yes i agree it is a little far away … i am going to be filming it anyway, so i will upload them to youtube and i will msg them to you
