invest in estonia

Menu
icon-small-x

The robot cleaning up in snowy Estonia’s AI sector

Developed after regular jobs ended for the day, in a whitewashed space optimistically named the ‘development centre’, this autonomous snow-clearing robot, now known as Lumebot, has gathered an impressive amount of traction over the last year. The company’s COO Kaspar Kikerpill addressed a packed-out room at Telliskivi’s Palo Alto Club on Wednesday 28th August, taking the curious crowd through their story, from the early days of the project to where it’s at today.

The Lumebot robot concept was born in August of 2018, inspired by a competition hosted by Cleveron, an Estonian robotic package delivery solution. Kikerpill, along with CEO Andres Kõiva, rose to the challenge of designing, building and entering a creation with their eye on thev 10,000 EUR prize.

The first attempt saw them take second place, but that was just the beginning of the road. The guys went on to develop Lumebot V2, a larger more capable machine that has seen them generate interest from the AI community and the media, as well as attracting a handful of paying customers – quite the achievement for such a fledgling company.

The brains of the operation

Under the bright red hood of the Lumebot V2 is a combination of cameras, radars and ultrasonic sensors that calculate, learn and manage the robot’s operation. Like the technology found in Starship’s delivery robots, Lumebot manages without the expensive Lidar sensors that significantly push up the price, the shortcoming of their only true autonomous US based competitor Left Hand Robotics.

Responsible for bringing the robots alive are the company’s team of six engineers, three of which are full time, the others working flat out alongside other jobs. As a forward-thinking nation, Estonia can struggle at times to find talent, AI specialists are highly sought after and therefore expensive. Lumebot are lucky to have a strong, dedicated team that have grown together.

What is AI anyway?

AI is an acronym for artificial intelligence and is often – and sometimes incorrectly – used to describe a range of autonomous processes, from something as rudimentary as a home robot vacuum cleaner to deep learning projects that aim to replicate free thought.

The process of enabling a system to populate its knowledge base through machine learning tools allows it a much deeper and far more complex understanding of a task or challenge than any developer would be able to programme in. It’s from this ability to collect and process data that an ‘intelligent’ system gains its advantage over one that has been preprogramed, and why such solutions are ideal for applications such as Lumebot’s, having the need to deal with unpredictable and ever-changing environments.

Where now for Lumebot?

The team state that by 2025 they will have 2000 robots clearing snow across the wintrier parts of the world. They project $50 million in annual revenue by then following a roll out that began in Estonia and will spread to the Scandinavian countries, before realising a wider goal of conquering the US and world markets.

They are currently warming up for a bid to raise a seed round and are looking for $1 million to help them take their team to nine and increasing their fleet of commercially capable robots to 15 by building a further 12, each worth $25,000 of revenue annually.

News & events

Need more information?

Need more information?

What is it like to run a business in Estonia? How to benefit from the e-solutions and the efficiency of our business culture? What are the opportunities in specific sectors?

The Estonian Investment Agency’s team is happy to help you via the complimentary e-consulting service.

Request e-consulting