These engineers are building robots to kill India's manual scavenging issue

While Genrobotic's 'Bandicoots' are capable of cleaning manholes in 20 minutes, the company is looking at using artificial intelligence for building fully autonomous robots

GenRobotics Bandicoot 2.0 in operation at Nellore municipal corporation
GenRobotics Bandicoot 2.0 in operation at Nellore municipal corporation
Peerzada Abrar Bengaluru
5 min read Last Updated : Mar 26 2019 | 11:24 AM IST
Jaya Singh is a contract worker with the Kumbakonam Municipality in Thanjavur district of Tamil Nadu. Not long ago, Singh and his other colleagues used to clean the manholes manually. But during this process, they were exposed to a hazardous environment as sewers were filled with poisonous gases, bacteria, viruses and parasites that cause various life-threatening infections among drainage workers. The lack of safety equipment such as ventilators, gloves, face masks and gas concentration detectors also put their lives at risk. Thanks to the robots created by Genrobotic Innovations, a Kerala-based startup, life is now much different for Singh and his colleagues. 

The remotely controlled robot ‘Bandicoot’ can go down into a manhole, spread its expandable limbs like a spider and scoop out the solid and liquid waste that blocks urban sewers. It has a robotic arm that can perform 360-degree motion and sweep the floor of the manhole to collect the waste in a bucket. The robot can clean a manhole in 20 minutes, compared with over two hours that at least three workers would take to do it manually.

“We are augmenting and improving the expertise of sanitation workers using these robots,” says Vimal Govind MK, co-founder and chief executive at Genrobotic Innovations. Apart from the social impact it creates, Govind says, the opportunity before the firm is huge as there are more than 4.5 million manual scavengers and over 60 million manholes in India. The market is also becoming big for the firm as infrastructure in the country grows.

An alumnus of the University of Calicut in Kerala, Govind co-founded Genrobotic in 2015 along with his college mates Arun George, Nikhil NP and Rashid Bin Abdulla. The co-founders, who were pursuing different streams of engineering like computer and mechanical, started Genrobotic as a college project to build an exoskeleton robotic suit for augmenting the capabilities of humans. 

After graduation, they joined various organisations to do regular jobs. “Initially it was exciting but we later got bored and decided to work on our college project separately,” says Govind. During the same time, a few manhole accidents took place in Kerala and the state government came up with a challenge to solve the issue by building an indigenous technology solution. The Genrobotic team decided to develop the robot to address this problem and launched the startup.

According to Genrobotic, its machines use powerful cameras and smart interface that can carry out the task with surgical precision. The simple user interface is helpful to the workers who cannot read. Its sensors detect and alert the presence of poisonous gases like ammonia, carbon monoxide, methane and hydrogen sulfide inside manholes. This could save workers from asphyxia. Also, the five-machine vision system of ‘Bandicoot’ gives a clear view of operations going inside the manhole in real time to the operator outside. The robots have been made rugged to withstand the harsh environment inside using materials like stainless steel and carbon fibre. The company is now working on making autonomous robots that can clean manholes and sewers using artificial intelligence.

GenRobotics Bandicoot 2.0
Genrobotic’s innovations received a huge stamp of approval when technology giant Google selected the startup for its ‘Launchpad Accelerator’ programme. “Startups like Genrobotic are a real inspiration for founders solving for India and its real needs,” says Paul Ravindranath G, programme manager, Launchpad Accelerator, India, Google. “Their ‘Bandicoot’ robot, with its semiautomatic spider robotic system for manhole cleaning, takes a serious shot at eliminating the practice of manual scavenging in India,” he adds.

Besides mentoring, the Launchpad programme has helped the firm connect with investors, including Rajan Anandan, vice-president, Southeast Asia and India, Google. Anandan has invested in the firm in his personal capacity.

The company has bagged projects from various state governments, including Kerala, Tamil Nadu and Haryana, as well as the Dubai Municipality in the UAE. It hopes to implement its technology across the country. “Apart from a deep social angle, Genrobotic is a stable and scalable business with a clear robotics-as-a-service delivery model,” says Bhaskar Majumdar, founder and managing partner at Unicorn India Ventures and an investor in Genrobotic.

The company said that it was also setting up training camps for manual scavengers to teach them how to use these robots in a safe and effective manner. It is also organising free medical camps and awareness classes for the community of manual scavengers as part of its effort to rehabilitate them. So far, Genrobotic’s team has trained more than 120 manual scavengers in different states.

Iron Man Suit

Genrobotic is also continuing its work on building exoskeleton robotic suits. This includes the 10-ft-high Generation one (G-1), in which the operator sits inside the suit. It can perform movements like a human with robotic legs and hands and can lift heavy objects with ease in rough terrains. It has applications in various areas like defence and space. 

GenRobotics Bandicoot 2.0
The firm's G2 product is a much more refined version of G-1 and uses pneumatic controls that give more efficiency in terms of power and utility. The arms of G2 are equipped to lift weights up to 60 kg and its legs can lift weights up to 160 kg, carry weights around 100 kg, and yet move around from one point to another with flexibility. The firm is now working on building artificial intelligence-based G3, which will be a wireless machine with an on-board compact power source and in-built voice recognition system with vision sensors and face recognition. It will be capable of performing almost all human activities.

Next Story