No announcement yet.

Advice for new teams - Torqeedo Thrusters

  • Filter
  • Time
  • Show
Clear All
new posts

  • Advice for new teams - Torqeedo Thrusters

    If you have attended RobotX before, you will already know what's contained below. If you're new to RobotX, then you will do well to read this post and learn from the experience of previous teams regarding propulsion systems. Mods, if you consider this post to be improper, feel free to take it down or contact me and I will edit it on request.

    When you look for propulsion systems for your WAM-V, you will come across a manufacturer called Torqeedo, whose products look wonderful. They're well priced, very powerful, and look like they will be perfect for your application. They even handle their own motor control and have batteries ready to go. Complete Torqeedo systems can be easy to set up and they will run like a dream. This is all true, but you should look past them for propelling an autonomous vessel.

    Torqeedo thrusters are excellent when used with Torqeedo batteries and a Torqeedo hand controller, but they are a perpetual nightmare if you attempt to use them by themselves. To the best of my knowledge, every team which has used Torqeedo thrusters at RobotX has experienced difficulties with them, and they have an unnervingly high failure rate at competition. The USYD and Georgia Tech teams at RobotX 2018 both had Torqeedo units fail within days of each other.

    Not only will they fail on you at the worst possible time, but the thrusters are also incredibly difficult to interface with from the beginning. Torqeedo will provide you with documents outlining the communication protocol that their thrusters use for control, so long as you first sign a Non-Disclosure Agreement (NDA). Once you sign the NDA and start trying to interface with you brand new thrusters, you will quickly discover that their documentation doesn't reflect the actual protocol in use, so you will still need to reverse-engineer the protocol yourself with only a small head-start over someone who didn't sign the NDA. If you think this process will be easy, it won't be. USYD spent months working on this task and troubleshooting our code and they still would randomly cut out on us whenever it would be most inconvenient for us. Even when they're running properly, they have an uncomfortably low slew rate which your controllers will not appreciate.

    Don't get me wrong, Torqeedo products are not all bad. Their battery units seem good, I haven't heard a complaint about them and they seem rugged and reliable. The failures, communication hardships and slew rate issues will also not cause problems for you if you mount them to a small boat which a human will be operating by hand, which is the original purpose of the hardware. USYD's 2018 hardware failure seems to have occurred when we gave the unit multiple abrupt command changes in a short time, so a hand-operated controller may have prevented that failure as well. The problems only arise when you attempt to control the Torqeedos directly through a computer, and that's the situation in which we find ourselves as people developing autonomous vessels. Torqeedo products are great at doing what they're designed to do, they're just not at all suitable for hacking. If you don't try to hack them, they will serve you well and I would definitely recommend them to the everyday consumer for their everyday boat.

    tl;dr Although they look really attractive and say they'll treat you well, don't get married to Torqeedo thrusters until you've found out why all their previous partners have divorced them.

    If any previous competitors have recommendations for good thrusters, please post them below. I personally think that the Minn Kota 80lb trolling motors are great so long as you don't mind going at a more relaxed speed and you have a few days to integrate them into your system. They're simple to work with and will run beautifully once you get them into the water.