View Single Post
  #30   Spotlight this post!  
Unread 02-02-2017, 20:16
Jared Russell's Avatar
Jared Russell Jared Russell is offline
Taking a year (mostly) off
FRC #0254 (The Cheesy Poofs), FRC #0341 (Miss Daisy)
Team Role: Engineer
 
Join Date: Nov 2002
Rookie Year: 2001
Location: San Francisco, CA
Posts: 3,082
Jared Russell has a reputation beyond reputeJared Russell has a reputation beyond reputeJared Russell has a reputation beyond reputeJared Russell has a reputation beyond reputeJared Russell has a reputation beyond reputeJared Russell has a reputation beyond reputeJared Russell has a reputation beyond reputeJared Russell has a reputation beyond reputeJared Russell has a reputation beyond reputeJared Russell has a reputation beyond reputeJared Russell has a reputation beyond repute
Re: What will be the next technical growth leap for the average team?

The general pattern for these things tends to be:

1. Innovation - Some enterprising team develops some technical innovation on their robot and has at least modest success with it (and often the most visible team to have success with something isn't the first team to try). Often this involves crossing some barrier to entry (overcoming manufacturing challenges, creative sourcing of a non-standard FRC material, advanced software expertise, lots of iteration, etc.) that keeps most other teams from being able to do the same thing immediately.

2. Early adoption - Other teams (whether nearby, of similar expertise, and/or related to the somehow through the diaspora of alumni and mentors) are inspired, and adopt a similar innovation and add their own twists to it. Whitepapers are released, code/CAD is open-sourced, etc.

3. Commoditization - Vendors take note...the AndyMarks, Vex's, West Coast Products, REVs, WPIlibs, etc., of the world begin to offer partial or complete solutions that significantly lower the barrier to entry for the average team. New companies spring up around the new products (Kauai Labs, CTRE, 221, etc). FRC "standards" emerge to allow for modularity in design and avoid vendor lock-in. Teams of "average" means and expertise can now utilize the idea.

(This is the standard lifecycle of adoption of a new idea...)


I've watched this all unfold time and time again. I can remember a time when each of the following had not yet attained ubiquity among "average" FRC teams...15 years ago there were entire regionals which would not have featured a single instance of:
  • 6 wheel dropped-center drive trains
  • #25 chain
  • 5mm HTD belts
  • Dog-shifting gearboxes
  • Modular planetary gearboxes (VersaPlanetary)
  • Wheels using incline conveyor belting (roughtop/wedgetop)
  • Omni wheels
  • Mecanum wheels
  • Swerve
  • 1.875" bolt circles for wheels, hubs, gears, and sprockets
  • Half inch hex
  • Quadrature encoders
  • Gyros/IMUs
  • PID controllers
  • Vision tracking of a stationary target using Rio or driver station laptop
  • Linear output, high-update rate speed controllers
  • CAN
  • Field-oriented driving
  • Sophisticated dashboards and telemetry

All of the above are now basically ubiquitous (and I'm sure there's a bunch more I missed); I can't remember the last time I saw a robot with none of these features.

To get back to the OP's question, I think there are a handful of ideas currently in the "Innovation" or "Early Adoption" phase. Yes, these are mostly on the software side (partially because I'm heavily biased, but also because FRC is far more mature mechanically than in software).
  • Multi- (more than 2) jointed arms
  • Motion profiling (this is on its way to ubiquity)
  • Offboard computing for "real-time" vision (ex. Tegra, Raspberry Pi, Android, etc.)...becoming more and more common.
  • Complex driving autonomous modes (2D path planning and trajectory following)
  • Depth sensing for object recognition (ex. Kinect or stereo vision)
  • Heads-up displays and "first person" driver control

...as well as a few capabilities that have not yet been widely featured on an FRC field, but might in the next couple of years. In most cases here, I think the real opportunity for innovation is from the commoditization of robotics technologies from consumer products and early-stage robotics industries ramping up capability while driving down cost:
  • 2D/3D Scanning LIDAR
  • Solid-state LIDAR
  • Torque control
  • Robot localization / SLAM - Precise drift-free robot pose estimation by recognizing landmarks on the field
  • Fast, accurate vision tracking of less structured targets
  • Brushless motors
  • Hydraulics and hybrid motor/fluid actuators
  • Low-cost, low-backlash gearing
Reply With Quote