# What DirectHop is
DirectHop is a tiny hopping robot developed at the University of Washington. The prototype weighs roughly 1 gram, fits in the palm of a hand, and can jump high enough to clear a standard stair step. Instead of using spring-and-latch hardware like many small bioinspired jumpers, DirectHop relies on a small electric motor to generate the launch impulse.
# How it controls jump height
The motor directly accelerates mass to produce lift. By adjusting the current sent to the motor, the team can tune the robot's jump height with single-centimeter precision. That approach avoids the binary "all-or-nothing" discharge of traditional spring-latch systems and simplifies miniaturization because it removes complex micro-scale springs and latches.
# Mechanical design and recovery
# What the prototype can and cannot do now
- Multi-hop operation: it can perform repeated jumps and ready itself for another hop after landing.
- Precise jump height control: centimeter-level adjustment of hop height via motor current.
- Self-righting: returns to a standing position about 90% of landings.
- No onboard sensing or navigation electronics yet.
# Planned improvements and research direction
The research team has a clear roadmap to address those gaps. Proposed upgrades mentioned by the authors include:
- Onboard power: small batteries or solar cells to remove tethered power.
- Steering mechanisms: a vibration motor to induce spin and tiny retractable feet to adjust launch angle.
- Onboard sensors: a camera and processing electronics to detect obstacles (for example, stairs) and compute needed jump height.
Their longer-term demonstration target is autonomous stair climbing: the robot would detect a stair, compute the required hop, position itself, jump, self-right, and repeat.
# Use cases and cost rationale
The team argues that component-level costs could be low enough to make inexpensive, semi-disposable DirectHop units practical. The vision is deploying large numbers—hundreds or thousands—of simple robots to perform distributed inspection or monitoring tasks where losing some units is acceptable. Suggested applications in the reporting include refinery inspection, agricultural monitoring, and planetary exploration, where hopping uses much less energy than continuous flight.
# Presentation and authors
The paper and prototype are scheduled to be presented on Sept 30 at the International Conference on Intelligent Robots and Systems (IROS). Principal contributors named in the coverage include Sawyer Fuller (senior author, UW associate professor of mechanical engineering), lead author Hanquan (John) Wang (graduate research assistant), and co-author Yash Talwekar (research assistant). The work was funded by the National Science Foundation.