Surveyor Robotics Journal |
|||||
|
Surveyor Corporation home page
Subscribe to a syndicated RSS feed. Archives |
Mon, 27 Feb 2006 new SRV-1 controller circuit board arrived today We got the new circuit boards, and have assembled the first one. We'll do functional tests tomorrow, but it looks pretty good so far. We're going to make a couple of small changes to the next batch of boards, but should still be on track for first robot shipments around March 10.
Posted Mon, 27 Feb 2006 21:51 | HTML Link | see additional stories ... Fri, 24 Feb 2006first test of motorized object tracking with SRV-1 Here's a quick test of a motorized object tracking function we're adding to the SRV-1. It's pretty basic, but works surprisingly well. Here's what the base station console looks like when we first sync to the target object (in this case, my hand).
Here's a short sequence of photos showing how the robot follows my hand.
Or click here to see a 5MB Quicktime movie of the sequence. It actually took less than 20 lines of C-code to implement this function in the SRV firmware. Clearly, there's a lot of opportunity to create better filters, smoother motion, etc, but this isn't a bad start ... Posted Fri, 24 Feb 2006 16:45 | HTML Link | see additional stories ... Mon, 20 Feb 2006new SRV-1 PCB layout and schematic We've had requests for information from users about interfacing the SRV-1 controller to other sensors, motors, etc. Here is the schematic and layout of the controller circuit board that just went into fabrication. Click on the images below to get higher resolution versions ...
Posted Mon, 20 Feb 2006 19:08 | HTML Link | see additional stories ... Sun, 19 Feb 2006new SRV-1 pricing; first shipments around March 10 The production circuit board design has been completed and submitted for fabrication. It appears that we will be able to start shipping robots around March 10, so the shopping cart will set up for orders at the beginning of March. We found some economies in the latest design which allow us to reduce the costs of the robots somewhat, so have revised the pricing. These are likely to be the finalized numbers. ======================== SRV-1 starter kit ....... $375 retail - includes one complete SRV-1 robot (controller and base), one SRV-1 base station, and one battery charger. $20-$55 savings from purchase of individual components ======================== SRV-1 robot ............... $275 retail - includes SRV-1 controller with ARM7 processor, digital video camera, 802.15.4 Zigbee radio, integrated with SRV-1 robot base SRV-1 base station ...... $95 retail - 802.15.4 Zigbee base station radio with USB interface, able to connect simultaneously to multiple SRV-1 controllers. includes SRV-1 Java console with built-in web server for remote access to SRV-1 controllers battery charger ........... $35 retail - Powerstream PST-5830-6 - 6-cell NiMH battery charger, wall mounted, charges at 1AH. 110VAC 60Hz source ======================== SRV-1 controller only .. $210 retail - includes ARM7 processor, digital video camera, 802.15.4 Zigbee radio, with motor control interface for 2 PWM DC motors. requires 7.2V power source SRV-1 robot base only ... $75 retail - 5" diameter anodized aluminum baseplate with 2 Solarbotics GM3 gearmotors, pivoting tail wheel, 6-cell 2AH NiMH battery pack, power switch, battery charger plug, 1-inch standoffs for mounting SRV-1 controller and camera ======================== Posted Sun, 19 Feb 2006 17:43 | HTML Link | see additional stories ... Sat, 18 Feb 2006SRV-1 object tracking ... first steps The next release of SRV-1 firmware will have basic object tracking functionality. A couple of buttons have been added to the base station console to capture the characteristics of an object and then try to locate it in subsequent frames. The process is simple and surprisingly robust. Here are some screen shots from initial tests:
The image on the left shows the live feed. Clicking the little green cross in the lower corner of the console cause the capture of object characteristics from a centered template (this will be moveable), and the image on the right shows the tracked object (in this case, my hand).
In this sequence, I caught Mickey's face. The template with two green crosses will actually capture 2 sets of object characterisics - in this case, we might want to get his face and his feet to more robustly track a specific object. Once the tracking is further refined, we'll hook up motor controls and instruct the robot to keep the target object in its field-of-view, or search around for an object matching the target characteristics. By the way, if you're wondering about the other console buttons, the little figures in yellow or brown are intended to turn on or off the motorized object tracking function. Posted Sat, 18 Feb 2006 18:13 | HTML Link | see additional stories ... new project - SRV-1 controller on an R/C tank base Here' s the tank with an SRV-1 controller installed.
We added a plug for the battery charger in the back.
The camera is temporarily attached to the front of the vehicle. When we have a chance to build driver circuits for the turret motors and airsoft gun, we'll move the camera up to the turret.
Posted Sat, 18 Feb 2006 16:59 | HTML Link | see additional stories ... Tue, 14 Feb 2006new project - SRV-1 controller on an R/C tank base Started surgery today, removing the existing R/C controller and sorting out the various signals. The interface task is pretty simple - there are two drive motors for tank tracks, and those will connect directly to the existing 2-channel PWM output from the SRV-1 controller. Â Those are the red and while wire pairs shown in this first photo. Â The other red, blue and black wires lead to the battery pack and power switch.
The gun turret interface will take a bit more work. On the left (black and white wires) is the turret pivot, and it's bidirectional, so we'll probably just build a simple h-bridge or else hook up a small relay to control this. The small black wire coming up through the turret connects to the radio antenna, and the larger assembly in the middle of the spocket is the airsoft gun with a tilt mechanism. The tilt mechanism is a simple cam, so the motor only operates in one direction. The red wire into the gun assembly is grounded, and if the black wire gets +voltage, the tilt cam starts to operate. If the black and white wire get +voltage, the firing mechanism is activated. So we just need to use 4 bits of digital output to control the turret swivel, the turret tilt, and the gun trigger.
Posted Tue, 14 Feb 2006 12:00 | HTML Link | see additional stories ... Mon, 13 Feb 2006new project - SRV-1 controller on an R/C tank base Look what UPS brought today ...
Wonder what they look like inside ?
Seems like there's plenty of room for an SRV-1 controller ...
Looks pretty simple to interface to the tank track drive motors. However, we'll need to add 2 H-bridge interfaces for the turret pan/tilt. Also, we need to see what kind of signal is required to fire the Airsoft gun. Probably a simple NPN transistor interface.
Check back to see how we're doing with the interface task, though we'll need to figure out where to mount the camera. Seems like it should go on the turret, though we might need to add some protection from hostile fire. Note that we have 2 of these things ... I wonder how we should use them ??? Posted Mon, 13 Feb 2006 12:00 | HTML Link | see additional stories ... Sun, 12 Feb 2006SRV-1 focus group at Little Scientist - Feb 11 These are some of the images of the focus group study workshop conducted on Sat Feb 11 and also in a small youth group on Feb 12.
Posted Sun, 12 Feb 2006 12:00 | HTML Link | see additional stories ... Thu, 09 Feb 2006SRV-1 focus group at Little Scientist - Feb 11 For those in the Toronto, Ontario, Canada area, there's an upcoming workshop on the SRV-1 robot at Little Scientist. See http://www.littlescientist.net/ for details. Posted Thu, 09 Feb 2006 13:55 | HTML Link | see additional stories ... Tue, 07 Feb 2006some preliminary photos and updated design notes We're getting ready to lay out the production version of the SRV-1 controller circuit board, and need to do some last testing of the additiions. Here's a version of the SRV-1 with 3 more IR emitters attached to a flying protoboard, so there are now a total of 4 emitters and 4 detectors.
We added small light shields around the IR emitters to filter out the light that was leaking out the back of the LED cases and directly into the IR detectors. Though the main purpose of the IR components is beacon detection and inter-robot communication, they work pretty well as proximity detectors, with a nominal range of 25-30cm (as long as the detectors aren't getting clobbered by stray light emission from the emitters, ergo the light shields). Posted Tue, 07 Feb 2006 12:00 | HTML Link | see additional stories ... Fri, 03 Feb 2006some preliminary photos and updated design notes The SRV-1 firmware updates can now be uploaded via the radio, and there's no more need for the RS-232 connector. We've made a plug that jumpers pins 4, 7 and 8 on a DB9 connector to control some signals that are used on the test robots to force the processor into flash bootstrap loader mode, but the production PCB won't use a DB-9 connector.
An explanation is provided here - http://www.surveyor.com/SRV_setup.html Posted Fri, 03 Feb 2006 12:00 | HTML Link | see additional stories ... |
||||