All projects

MACS: Magnetic AGV Contactless-charging System

Mar 2026 – Aug 2026 · Team Project · Hardware Lead

Indoor logistics AGVs usually stop working and drive to a charging station. That forces fleets to keep spare robots, manage battery state of charge carefully, and queue at stations. MACS changes where charging happens: commercial 82–85 kHz wireless power modules are installed beneath the waypoints where AGVs already stop to load and unload, so charging happens during work rather than in place of it.

Charging energy retained versus full-path charging on the same route
95.7%
Transmitter infrastructure cost (₩360,000 → ₩60,000)
−83.3%
Best measured end-to-end WPT efficiency
85.72%
Warehouse illustration with AGVs charging over wireless transmitter pads at waypoints
MACS concept: transmitter pads are placed only at waypoints where AGVs already stop to work.

Prototype

The team rebuilt a ROBOTIS TurtleBot into a wireless-charging AGV and demonstrated it on a 1200 × 600 mm physical track. A tag-recognition camera, a CC/CV charging module, a battery pack, and the wireless receiver were integrated on the robot; an AprilTag-based alignment routine positions the receiver coil over the transmitter at each waypoint.

Modified TurtleBot with tag camera, CC/CV module, battery pack, and Tx/Rx wireless charging coils
Prototype AGV: tag-recognition camera, CC/CV module, battery pack, wireless Tx/Rx coils, and an AprilTag for coil alignment (labels in Korean).

System architecture

On the station side, a 220 V supply feeds a 100 W, 12 V SMPS, a relay, and a proximity sensor ahead of the wireless transmitter. On the vehicle, a Raspberry Pi 5 main controller coordinates a Raspberry Pi 4 for the TurtleBot powertrain, a second Raspberry Pi 5 for the manipulator, and a cargo-bed actuator; an 18650 3S1P 12 V pack charges through the wireless receiver or a wired charging module.

Wireless AGV system block diagram with power paths in red and signal paths in blue
System block diagram. Red lines carry power; blue lines carry signals.

Validation

Transmitter input and receiver output power were measured with an oscilloscope and a bench power supply. End-to-end efficiency reached 79.96% at full coil alignment with a 5 mm gap and 85.72% at the best tested operating point.

On the same test route, waypoint charging delivered 0.8333 Wh versus 0.8704 Wh for charging along the entire path. Coils along the driving path contributed only about 4.4% of the energy, which supports installing transmitters only at waypoints.

Test bench with oscilloscope, power supply, TurtleBot, and wireless charging module
Efficiency test bench: oscilloscope, power supply, TurtleBot, and wireless charging module (labels partly in Korean).

Comparison with existing approaches

Wired chargingFull-path inductive chargingMACS
Where it chargesDrives to a stationAlong the entire pathAt work waypoints
InfrastructureSeparate stationsCoils along every pathSelected waypoints only
Work and chargingSeparatedCharges while movingSimultaneous
AdoptionNeeds station spaceLarge-scale constructionPhased, on existing routes

My role

  • Led hardware for a team of four: integrated the wireless transmitter and receiver into the TurtleBot and redesigned its battery and electrical system.
  • Proposed waypoint charging in place of a fully embedded track and designed the corresponding driving logic.

Award certificates

Certificate of the President of Korea Electrical Engineers Association Award, open category, 23rd IE Competition, dated July 14, 2026
President of Korea Electrical Engineers Association Award · 2026 KIPE IE Competition · Jul 14, 2026

Resources

Technologies: WPT · Raspberry Pi · ROS · Arduino · Fusion 360 · Isaac Sim