Team NEXTBOTS Develops Dog-Bot For Surveillance Across Varied Terrains

Team NEXTBOTS Develops Dog-Bot For Surveillance Across Varied Terrains

Students' Corner May 2020 Team MEXTBOTS Dog-Bot Surveillance Varied Terrains
{data-social}

Staying vigil at all times and minimising any loss of life has been among the top priorities for our defence forces, especially for a country like India that is home to an extremely varied terrain

NEXTBOTS, a six-member team from Sharda University has developed an animal-shaped robot named Dog-Bot that it claims can serve as the eyes and ears of our defence forces in ensuring regional security.

The automobile sector has always been at the forefront in helping other industries excel. Even as the industry has wholeheartedly come forward to support the fight against Covid-19 manufacturing medical equipment, other sectors such as defence have also relied heavily on the automotive industry for multi-role vehicles needed for combat operations as well as manage quicker troop movement, thereby maintaining higher security across the country. However, security forces often have to cope with challenging terrain such as Siachen Glacier that often makes their mobility cumbersome. Aiming at addressing such problems of mobility, NEXTBOTS, a six-member team from Sharda University, brainstormed on developing Dog-Bot – a quadruped animal-shaped robot that can aid tasks such as surveillance, defence as well as assault along with other applications.

(L-R) Ajay Kumar, Gowatham Podal, Manthan Gandhi
(L-R) Shashank Shekher, Vasu Dev and Vineet Prajapati from Team NEXTBOTS have brainstormed to develop the Dog-Bot that can serve as the eyes and ears of our defence forces in ensuring regional security

PROBLEM STATEMENT

Defence forces serve as the lifeline of any country as their vigilant eyes allow our countrymen to sleep at peace in their homes. Banking on a plethora of vehicles such as wheeled or tracked design, defence forces are able to navigate through a variety of terrains. There are times when defence forces need to respond to any unscrupulous activities or patrol remote areas that require robust planning and equipment for quicker problem addressal. At this stage, the major problem with wheeled-vehicles is that they need a track or a flat surface to work. Track vehicles have low top speed, much higher complexity and can damage the surface they move upon. Thus, there is the requirement for a solution that can quickly adapt to changes, work robustly in varied temperature and terrain as well as eliminate any loss of human life.

INSPIRATION

NEXTBOTS comprises a team of six B.Tech students of Sharda University handling different roles – Ajay Kumar handles hardware and software integration & Robot Operating System (ROS) programming; Gowatham Podal is responsible for ROS programming; Shashank Shekher for CAD-based structuring of various components & designing electronics hardware like Printed Circuit Boards (PCB); Vineet Prajapati for CAD design of various components); Manthan Gandhi handles marketing and logistics, while Vasu Dev handles designing parts.

The NEXTBOTS team derived the inspiration to build this robot from a problem statement of a competition named DRUSE (Defence Research & Development Organisation – Robotics and Unmanned Systems Exposition), a national-level robotics competition. The competition that was launched in 2017 is an open platform to synergise and popularise national-level talent in the area of robotics among the student community around development of unmanned systems that may find application in defence applications. Mountain climbing robots such as the Dog-Bot need to have robust characteristics to carry load as well as serve as the eyes of the soldier without leaving any track marks. Apart from various applications like defence surveillance to assault, search and rescue, the Dog-Bot can also be deployed across domestic applications such as home surveillance & security, industries applications & disaster management services.

SOLUTION

Dogs are autonomous, free-moving and intelligent creatures. The design philosophy of the Dog-Bot was to mimic an actual dog as closely as possible, while bearing electrical and mechanical limitations in mind. The Dog-Bot is a mechanically robust design that is currently at a prototype stage. For now, the team is using 3D print parts, which will soon be replaced with aerospace-grade aluminum computer numerical control (CNC) parts and carbon fibre as well as a custom-designed motor sourced from China to remain lightweight and survive high-impact falls or accidents.

The animal-shaped robot spans 30 cm x 27 cm x 30 cm in length, breadth and height respectively and weighs just 13 kg for agility. Large range of motion at each joint allows operation forwards, backwards, or upside-down and high speed, high torque actuators that let the robot perform omni-directional movement with trotting, pronking, bounding, and pacing gaits. The Dog-Bot is enabled movement by a custom-built modular, low-gear ratio, back-driveable actuators with integrated electronics and high-bandwidth torque control and can move 2.45 m/s forwards, 1 m/s sideways, in-place turning at 5 radians/sec.

The Dog-Bot is powered by a 120 Wh lithium-ion battery that can be fully charged in 70 min for a work range of 30-90 min, depending on the tasks executed. The brain of the Dog-Bot is UP Board with quad-core Intel Atom processor for locomotion control and STM32F4 microcontroller by STMircoelectronics in each actuator for joint-level control. The new generation of rough-terrain robots captures the best features of animals, like their mobility, anatomy and speed and can navigate various kinds of terrain including steep, rocky, wet, snowy, muddy, etc, claims NEXTBOTS. The Dog-Bot can be equally proficient in navigation across city streets as well as working around the usual household clutter, doing its chores, climbing stairs, etc., tasks, where wheeled vehicles have limited capabilities.

The team has been guided by various department heads at Sharda University, including Prof (Dr) Parmanand Astya, Dean of School Engineering & Technology; Dr Pallavi Gupta, Associate Professor, Department of Electronics & Communication and Prof (Dr) Ashok Kumar, Department of Physics. The team had to overcome numerous challenges in executing this project. The project took almost two months in designing the first prototype leg. NEXTBOTS used plywood for making its parts and used linear actuator, which was working but lacked the desired speed and payload capacity.

Further, the team made changes and design improvements that took almost a year, as linear actuators were replaced with high torque brushless drivetrains having sufficient power in a compact form factor. The Sharda University team invested around ₹ 600,000 and around three years to reach its current testing phase.

ROUND-UP

The evolution of Dog-Bot has equipped humans and animals with much better agility to navigate complex terrains. Robots having these capabilities would require terrain sensors, advanced actuators, power systems and sophisticated computing with dynamic controls. A technologically advanced quadruped robot that contains much of the advantages of the mobility of a real dog and can bear the load like a donkey would be very useful.

The team had qualified for the first round of DRUSE, where 30 teams were selected from among 1,300 submissions across the country. Although the team has developed its first prototype as a manual-driven solution, it is brainstorming currently to graduate it to an autonomous solution.

TEXT: Anirudh Raheja