Ladimo develops smarter machine vision
Autonomous cars must be able to detect pedestrians crossing the street or the lurching movements of other cars stuck in traffic. Assembly robots and augmented reality features in mobile applications also require accurate information regarding the shapes and distances of their environment. These are examples of cases in which the innovation of Ladimo, a company that originated from Aalto University, could serve as the eyes of a machine.
The Ladimo team develops machine vision based on 3D laser measurement – and not just any kind: 'Our 3D point cloud sensor is the most accurate real-time sensor in the world,' say Ladimo's CEO Jouni Halme and CTO Jorma Palmén.
Some partners initially expressed doubts about the claim, but Ladimo seems to have convinced them – thus far, the technology has attracted interest from a manufacturer of mining machinery, three car manufacturers and several international mobile device companies
An innovation born on a Sunday afternoon
The story of Ladimo begins in 2012 with an epiphany.
'The thought came to me by surprise on a Sunday afternoon, just as I was walking down the stairs in my home. As I sat down for coffee, I told my wife that I had just come up with an invention, and that this one might turn out to be major,' says Palmén, who at the time was employed as a lecturer of engineering geology and mineralogy at Aalto University.
On Monday morning, Palmén prepared an invention disclosure. With funding from Aalto University and the Finnish Funding Agency for Technology and Innovation Tekes (now Business Finland), he began developing a prototype in collaboration with Jussi Leveinen, a professor of geology, and Matti Kurkela, a researcher specialised in 3D measurement.
In simple terms, Palmén's invention led to a new, highly accurate method for measuring an object's distance from the sensor of a camera.
Ladimo's machine vision technology is based on a machine-mounted camera that detects a laser beam emitted by the machine and dispersed into thousands of points. With a measurement based on triangulation, the system calculates coordinates for every one of those points. The result is a 3D point cloud.
However, light beams bend as they move through the air, which causes errors in measurements. The optical elements of the camera also cause distortion. Earlier methods tended to split the laser beam into a random set of points, but Palmén's idea was to measure location of the points with a regular point set. With the help of modelling, this new method can correct the optical and geometric errors in measurements and, when necessary, remove points that are too distorted.
Once the idea was proven functional, Ladimo received funding from Tekes (Business Finland) to begin commercialisation. The company was founded in 2015. At that point, Jouni Halme joined the team. Halme has built an international career in sales and marketing that includes work for Nokia and various start-up companies.
Based in Otaniemi, Espoo, Ladimo's development team is currently composed of seven employees. The company's advisors include professors from Aalto University, the University of Eastern Finland and the Optoelectronics Research Centre of the Tampere University of Technology.
The right time for vision
Ladimo works in a field that is growing at an explosive rate – autonomous vehicles, virtual reality and augmented reality as well as robotics are quickly becoming everyday concepts. This is also evident in the hundreds of millions in funding invested in companies working on machine vision.
This year will see Ladimo introduce its first product for especially for robotics. At the same time, development is rapidly accelerating on smaller sensors fit for consumer devices. The company is currently working on a sensor for mobile phones and robotics applications, weighing in at only 60 grams.
The company is preparing to launch a round of funding where it seeks 2.5 million euros from strategic partners to develop the miniaturised sensor. Interest from large international technology companies has set the team's expectations high.
'We will exceed a hundred million euros in revenue within five years,' Halme and Palmén estimate when asked about Ladimo's future.
In this series we present research-based startups from Aalto University. Each year, 5 to 10 research-based companies are founded with the university's support.