Interview with C-Link Systems William Lovell

Interview with C-Link Systems William Lovell

C-Link Systems has been delivering a range of robots for over a decade. Some operate in close proximity with people. I talked with CEO and Senior Engineer, William Lovell, about this and their latest robotic arm, the C-Link Systems M18 (Fig. 1).

The M18 can be used in a stand alone configuration or mounted atop C-Link System's mobile Forager platform (see Forager Robot Has Tower Inside) that can utilize Freescale's Tower development boards.

Figure 1. C-Link Systems’ M18 robotic arm is designed for educational as well as industrial use.

Wong: C-Link Systems provides a range of products and services. Can you give us a quick overview and talk a bit more about your robotic arm?

Lovell:c-Link Systems was founded in February of 2000. I had been working in factory automation upgrades/retrofits with a specialization in fiber optics since the mid 80’s. Hence the lower case “c” in the name is the symbol for the speed of light. In 2005 Hurricane Katrina struck New Orleans, I received multiple emails and telephone calls asking if I had completed a pet project and if not “get it together”. As a company we had been slowly moving out of automation upgrades, it is a cut-throat business where bottom line seemed to be the only concern. From that disaster was born a new direction for c-Link and one I enjoy more. We now have commercial products; Forager-ARV and Vole-Bot, education products; M18e-Arm and Trak-Bot. In late 2011 I was asked to teach robotics in a local high school. I realized that the robotic arms available to high schools, community colleges and even some middle schools were low cost made, non-standard design or priced far out of reach by most. The M-18e was soon developed with a caveat that we were getting requests to do commercial installs. So the M-18i joined the M-18e, with some slight variations.

Wong: What makes your system differ from conventional robotic arms?

Lovell:We designed it. Actually a number or small things; the whole unit is aluminum with the exception of the slew rings and hardware, all major joints have slewing rings, we use stepper motors (allows for microstepping), limit points and hard stops on all major joints and on the commercial version there is position feedback (optional on the educational model). The physical unit is more in-line with an industrial robot like a Kuka KR-16 or ABB IRB-1600. The control system is predominately off-the-shelf and where we could reduce costing the control card was designed and fabricated in-house, most of those are also available. Our sell price is our best feature and in the case of education the “bang for the buck” is even higher. Robotic arms in education arena that come near our price are built with folded sheet metal (viable method for some things), plastic or a combination. The motors on those systems are in most cases RC model servos or sometimes gearhead DC motors, a few have feedback loops.

Wong: How does someone program the M18-Arm?

Lovell:Was wondering when we would get to this. My first thought is you are referring to application level programming. We will have four methods:

1.    The stereo-typical write your own code, compile and store in application memory. This requires an in-depth knowledge of motion, the arm and actual motion task. This is great for college level robotic engineers. The software for this is downloadable from freescale Semiconductor and is the compiler for the processor we use.

2.    We are going to employ an old method, I first used this on a Binks painting robot, lead the robot by the nose. Basically you move the end-effector through the motion you wish to take from start to finish, including operation of the end-effector. The whole time this is being done the feedback units are being recorded; hence the reason for feedback even on stepper motors. The action is than played back as a motion. Great for single point repetitive operations.

3.    This is similar to lead the nose but instead you use a teaching pendant. Here you enter position values for each segment to move too. After completion it is played back combining the moves on all axes into one motion.

4.    Use of the Microsoft Robotics Studio is the last part tackle. The cross between programming and graphics appeals to younger engineers these days. Basically you drag and drop different control and action blocks interconnect them and the need values such as speed and angular distance and the whole picture is compiled. There is a simulator section which is very handy to check your work. There is a lot more going on behind the curtain so this is currently being evaluated as to implementation method and time frame.

Wong: Is it safe to work around M18-Arm?

Lovell: We like to think so. Safety depth is also dependent on what the end-effector is. Machine shop safety practices should be used. So long as safety features like the limit points and hard stops have not been tampered with, there should be no issue. Common sense still has to play a factor, moving/rotating parts means keep your fingers out. You need to know what the envelope of motion is and not stand in it. For those that do the system monitors motor current, if it exceeds the set point the robot is shut down and warning will appear on the control system LCD. A future enhancement will be the use of sensors at the end-effector, wrist and elbow joint that actually “watch” for collision potential then slow or stop the arm prior. We are using this type of technology on our large AGV when the arms are in motion. We found humans become fascinated with watching usually followed by getting in the motion path “for a better view”.

Wong: What are some of the features of M18-Arm?

Lovell:As we were saying earlier; all aluminum, micro stepping motors, position feedback, minimum of 5-DoF, slewing ring joints, electronics is open sourced or COTS, low voltage motors, can be disassembled, universal mounting for end-effector and available I/O and motor control for an end-effector.

Wong: Where do you expect these robotic arms to be used?

Lovell: The intended use was for education to show/teach multi-joint arm motion. I had even gone as far as to create a lab for arm dynamics and the interaction with the joint motor. In the July of 2012 Robot Magazine the editor did a product review of the M-18 proto-type for education. We had a couple of schools inquire but our biggest hit was small shops looking for a low cost arm to do repetitive operations. One shop wanted to load/unload a small CNC lathe moving the machined part to a second machine for further machining. A current project is using an M-18 to load components into a holding jig for a glue process. The customer actually wants a number of M-18s doing all the mundane work on the jigs. So I would say we are seeing commercial as a more viable use area than education.

Wong: Does the system have any built-in intelligence or control?

Lovell: Yes the system is intelligent enough to run as a standalone without a host. A recent upgrade reduced our processor count but has improved coordinated motion. The core processor is a freescale PXD1010, uses a Power Architecture e200z0, this will be in the M-18e whereas the M-18i will have a PXD2020 (e200z4). The master processor will be a MCF52259 or MPC5125 which will handle Ethernet, serial ports, USB and other I/O functions. The choice of the processor is dependent on user needs and system level functions. Because we use the freescale Tower System users can add their own card to the Tower if there is a slot open.

Wong: How is M18-Arm linked to a host computer?

Lovell: The arm does not require a host system but if you are running a host system with a network card or access to a network than that’s it. We are going to include an Ethernet port into the master processor. This will allow for compiled application code to be downloaded or in lieu of a pendant use of the host computer. Currently the host will be a PC. Wi-Fi is being discussed but with all the other work going on it will be awhile before I move on that aspect. If your reference to link is between the physical arm and its electronics, we use flexible cable and low voltages in comparison to the commercial units.

Wong: How do you debug M18-Arm's applications?

Lovell: Currently will only be available at the code development level. We use the built-in debug section of CodeWarrior, which is the software utilized by freescale Semiconductor.

Wong: What are some new features we might see in M18-Arm in the future?

Lovell: Should I claim spoiler alert? First one out of the gate will be a MEMS based position feedback; this will be followed by the safety position sphere monitor. The largest optional item will be UGV-18; this will be a 4-wheel drive platform utilizing mecanum styled wheels. The UGV-18 will have the ability to mount and operate an M-18i via a Wi-Fi connection. Other add-ons will be in the end-effector arena: a palletizing bolt on system, milling/drilling head, we are having internal talks on a MIG welding head or laser cutter. Some of the end-effectors are driven by us for in-house use. Some of the add-ons may not be available to educators at the introduction time. We are also looking at Foragers different payloads to see what end-effectors could be scaled down for use.

Hide comments


  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.