Rover Profile/ 2019

Mongol-Tori2019

Fourth-generation Mars rover, rebuilt for retrieval, servicing, autonomy and science

Drive: Four-wheel modified rocker-bogie suspension
Arm: 6-DOF
Payload: Heavy claw lifts more than 5 kg
Autonomy: GPS + magnetometer + YOLOv2 vision (81% accuracy)

CompetitionLEAD // Razin Bin Issa

Scroll

Mission Brief

The 2019 BRACU Mongol-Tori rover marks the fourth iteration in the team's URC lineup, a four-wheeled rover built to assist astronauts and run autonomous science research on Mars-analog terrain. This year's build focuses its major upgrades on the software system, electronics, and a fully automatic on-board science laboratory, with supporting refinements to the chassis, arm, and wheels. It pairs an expandable modified rocker-bogie suspension and a 6-DOF arm with three interchangeable claws against a feedback-driven control GUI and a computer-vision autonomy stack.The 2019 BRACU Mongol-Tori rover marks the fourth iteration in the team's URC lineup, a four-wheeled rover built to assist astronauts and run autonomous science research on Mars-analog terrain. This year's build focuses its major upgrades on the software system, electronics, and a fully automatic on-board science laboratory, with supporting refinements to the chassis, arm, and wheels. It pairs an expandable modified rocker-bogie suspension and a 6-DOF arm with three interchangeable claws against a feedback-driven control GUI and a computer-vision autonomy stack.

Spec SheetDWG mongol-tori-2019-2019

Competition: URC 2019
Year: 2019
Team Lead: Razin Bin Issa

01 / What Makes It New

5 breakthroughs that define Mongol-Tori 2019

01
Innovation
Expandable rocker-bogie suspension
A four-wheel modified rocker-bogie whose bogies extend from 0.76 m to 1.2 m via two heavy-duty linear actuators, giving the rover added stability through vertical drops and steep slopes.
02
Innovation
Feedback-driven arm control
A new feedback system integrates feedback end effectors into the Java control GUI and replaces camera-based arm decisions with angle calculation, improving precision for equipment servicing.
03
Innovation
Automatic on-board science laboratory
A two-box, five-test lab (biomass, amino acid, water flow capillary, spectroscopy, microscopy up to 800x) fed by a fully 3D-printed central soil distribution system, turning the rover into an automatic planet analyzer.
04
Innovation
Computer-vision autonomy stack
GPS and magnetometer fused with darkflow/TensorFlow/OpenCV and YOLO v2 vision (81% accuracy) on a custom OpenStreetMap-referenced offline map to autonomously detect and reach targets.
05
Innovation
Tracking communication link
Dual Rocket M2 / Bullet M2 routers with a G-5500 rotator and elevation-azimuth controller that auto-tracks the rover by signal strength, plus separated 5.8 GHz video and 2.4 GHz control bands.

02 / Engineering

Built subsystem by subsystem

Every discipline on the team owns a slice of the machine. Here is how each one comes together.

SYS.01Mechanical

01Subsystem

Mechanical

A reworked mechanical platform lowers the rover's center of gravity for stability while keeping the four-wheel layout. The K-shape space-frame chassis, expandable rocker-bogie suspension, and redesigned stainless-steel wheels are tuned for rough terrain, vertical drops, and slopes.

K-shape chassis 0.7 m x 0.55 m with triangulation to resist shock-induced distortion
Modified four-wheel rocker-bogie with two bogies and a U-shaped differential bar
Universal joints in the differential bar maximize bogie rotation angle
Bogies expand from 0.76 m to 1.2 m via two heavy-duty linear actuators for drop stability
0.3 m x 0.10 m stainless-steel wheels with U-shaped rim, drilled holes, and rubber grip pads
6-DOF arm: 4 linear actuators, 2 DC motors, motor-and-pulley base, gear mechanism to avoid wire twisting

SYS.02Network

02Subsystem

Communication

A dual-router link carries video and control between rover and a portable base station, with antenna tracking to maintain coverage. Multiple camera feeds and frequency separation keep video and control reliable.

Rocket M2 at base station and Bullet M2 on the rover
Omnidirectional TL-ANT2412D on rover; high-gain directional ANT232D15T-120DP at base
G-5500 rotator with elevation-azimuth controller tracks the rover by signal strength
3 IP cameras + 6 FPV cameras, viewed on 3 monitors and 1 VR set
UDP for fast video, TCP for reliable control signals
5.8 GHz band for FPV video; 2.4 GHz for control signals and IP cameras

SYS.03Electronics

03Subsystem

Electronics

The electronics sector received a major upgrade with new components selected from prior testing. High-torque geared motors, multiple Cytron drivers, and a layered battery scheme power drive, arm, science, and comms, with kill-switch and fuse safety.

High-torque DC geared motors; rated/avg/stall current 10/14/20 A
Four 30 A Cytron MD30C drivers for wheels, six 10 A Cytron MD10C drivers for arm
ATMega 2560 controls wheels and arm; PCBs designed in Proteus and Eagle
10,000 mAh Li-Po (~42 min); 2 Li-Po packs for wheels; 24 V Li-Po for router/NUC/switch
Science power: two (1p2s) Li-ion and one (2p3s) Li-ion packs
High-ampere kill switch plus a 30 A fuse box for the main circuit board

SYS.04Controls

04Subsystem

Controls & Software

Multiple GUIs handle control, science, and mapping over a TCP client-server framework. A new feedback system is a key upgrade, letting feedback end effectors drive arm precision.

Java control GUI; base station is client, rover is server over TCP
Feedback-receiving system in the control GUI for feedback end effectors
C# science GUI translates sensor readings into custom measures with graphical recognition
Mapping GUI tracks the rover on a pre-loaded offline map via GPS lat/long
Platform-independent control GUI compatible with any system and rover

SYS.05Autonomous

05Subsystem

Autonomy

The autonomous stack fuses GPS, magnetometer, and camera vision to traverse to a target. GPS guides the rover to a radius of the goal, then an image-processing routine detects the marker and homes in.

Sparkfun Venus GPS (Venus638FLPx) and HMC6343 three-axis magneto-resistive module
GPS gives lat/long with ~5 m error against a preloaded map
Custom offline map built in Python with geo-referencing from OpenStreetMap
Within 3.5 m of destination the image-processing sub-routine takes over
Vision built on darkflow, TensorFlow, OpenCV, and YOLO v2 at 81% accuracy to detect the tennis ball
C++ autonomy software combines GPS, compass, and camera for full autonomy

SYS.06Science

06Subsystem

Science

The science module turns the rover into an automatic planet analyzer across three segments: a digging sample collector, an environmental sensor box, and an automatic on-board laboratory. It runs five precise tests across two boxes to search for biosignatures and characterize soil.

Excavator claw on the 6-DOF arm digs hard soil up to 20 cm, supplying up to 80 g per collection
Sensor box: DHT22 (air temp/humidity), DS18B20 (soil temp/moisture), MQ7 CO, MQ135 CO2, MQ8 H2, Grove O2
More sensors: LDR light, ML8511 UV/radiation, Hall effect, compass, barometric, anemometer, pH probe, endoscopic camera
Box 1 runs Biomass (load cell + nichrome heater), Amino Acid (ninhydrin reagent), and Water Flow Capillary (syringe + servo) tests
Box 2 runs Spectroscopy (N, P, K quantitative analysis) and Microscopy (digital microscope up to 800x)
Fully 3D-printed central soil distribution system feeds each box automatically

03 / Telemetry

The numbers behind the build

Chassis

0.7 m long x 0.55 m wide K-shape space frame

Drive System

Four-wheel modified rocker-bogie suspension

Suspension

Expandable bogies, 0.76 m to 1.2 m via 2 linear actuators

Wheels

0.3 m dia x 0.10 m width, stainless steel, rubber-padded

Arm DOF

6-DOF with 3 interchangeable claws

Payload

Heavy claw lifts more than 5 kg

Power

10,000 mAh Li-Po, approx. 42 min runtime

Compute

NUC mini PC + ATMega 2560

Range

Designed to travel up to 1 km

Autonomy

GPS + magnetometer + YOLOv2 vision (81% accuracy)

Parts Index // 20 components

NUC mini PC
ATMega 2560
Cytron MD30C
Cytron MD10C
Rocket M2
Bullet M2
TL-ANT2412D antenna
ANT232D15T-120DP antenna
G-5500 rotator
Sparkfun Venus GPS (Venus638FLPx)
HMC6343 magnetometer
YOLO v2
TensorFlow
OpenCV
darkflow
DHT22
DS18B20
MQ7 / MQ135 / MQ8
ML8511 UV sensor
Proteus / Eagle
NUC mini PC
ATMega 2560
Cytron MD30C
Cytron MD10C
Rocket M2
Bullet M2
TL-ANT2412D antenna
ANT232D15T-120DP antenna
G-5500 rotator
Sparkfun Venus GPS (Venus638FLPx)
HMC6343 magnetometer
YOLO v2
TensorFlow
OpenCV
darkflow
DHT22
DS18B20
MQ7 / MQ135 / MQ8
ML8511 UV sensor
Proteus / Eagle

Proteus / Eagle
ML8511 UV sensor
MQ7 / MQ135 / MQ8
DS18B20
DHT22
darkflow
OpenCV
TensorFlow
YOLO v2
HMC6343 magnetometer
Sparkfun Venus GPS (Venus638FLPx)
G-5500 rotator
ANT232D15T-120DP antenna
TL-ANT2412D antenna
Bullet M2
Rocket M2
Cytron MD10C
Cytron MD30C
ATMega 2560
NUC mini PC
Proteus / Eagle
ML8511 UV sensor
MQ7 / MQ135 / MQ8
DS18B20
DHT22
darkflow
OpenCV
TensorFlow
YOLO v2
HMC6343 magnetometer
Sparkfun Venus GPS (Venus638FLPx)
G-5500 rotator
ANT232D15T-120DP antenna
TL-ANT2412D antenna
Bullet M2
Rocket M2
Cytron MD10C
Cytron MD30C
ATMega 2560
NUC mini PC

04 / Mission Plan

Four missions, one machine

How this rover is engineered to score across every University Rover Challenge task.

01
Extreme Retrieval and Delivery Mission
The rover is designed to travel up to 1 km, with the modified rocker-bogie suspension absorbing sudden shock on rough terrain and the expandable body landing firmly through vertical drops. The heavy claw lifts heavy items such as tool boxes or water bottles.
02
Equipment Servicing Mission
A 3-finger claw and feedback end effectors deliver the precise, strong movement needed to operate switches and keyboards. The team shifted from camera-based arm decisions to angle calculation, determining both the movement that occurred and the movement still required.
03
Autonomous Traversal Mission
Fusing Venus GPS, an HMC6343 magnetometer, and camera vision, the rover navigates a preloaded map to within 3.5 m of the target, then a YOLO v2 / OpenCV sub-routine detects the tennis ball and computes the path. The magnetometer keeps the rover aligned to the determined heading.
04
Science Mission
An excavator claw digs and collects soil for an automatic on-board laboratory that runs five tests (biomass, amino acid, water flow capillary, spectroscopy, microscopy) alongside a multi-sensor environmental box to assess biosignatures, climate, and soil composition.