Seeing the Future through the Lens of Computer Vision: our investment in Robovision
The emergence of Generative AI has taken the tech world by storm, heralding what many perceive as a significant leap forward in artificial intelligence capabilities.
But Gen AI isn’t the sole avenue for creating meaningful value through AI. There are other areas that offer substantial potential for value creation. For instance, applications in healthcare, agriculture, production, education have demonstrated tangible benefits.
For many of these applications, Computer Vision — which focuses on enabling machines to interpret and understand visual information from the surrounding environment, much like the human visual system — is the underlying catalyzer.
Computer vision encompasses a broad range of techniques and algorithms designed to analyze, process, and extract meaningful insights from images and videos.
- Object detection and recognition
- Facial recognition
- Optical Character Recognition (OCR)
- Gesture recognition
- Scene understanding
- Sentiment analysis from faces and gestures etc…
Most of the application are in 2D so could work with only one camera, but some of them would require 3D Computer Vision systems.
The expansion of computer vision use cases is fueled by Staff shortage.
The issue of staff shortage in agriculture and manufacturing sectors has been a pressing concern since COVID19. I live in Geneva where each year, I hear winemakers saying finding seasonal workers to pick grapes is more complicated than the previous one.
This a global phenomenon: in many countries, the demand for skilled workers far exceeds the available labor pool. According to recent data by the US Chamber of commerce, there are 9.5 million job openings in the U.S., but only 6.5 million unemployed workers. Most of the shortage is in the agricultural sector, leading to significant challenges in planting, harvesting, and maintaining crops. But similarly, in manufacturing, the shortage of skilled labor is a critical issue. Studies indicate that by 2030, there could be a global deficit of approximately 2 million manufacturing workers.
Furthermore, in many countries, policies aiming at reducing the stream of immigration have unfortunately rendered the issue even more pressing.
Computer vision combined with robotics can mitigate this labor shortage and is already doing so in many areas and sectors.
In addition, it holds the potential to significantly reduce the exposure to risk and pollution for workers in hazardous environments.
This can be done thanks to
- remote monitoring: avoiding workers to physically enter hazardous areas such as chemical plants or construction sites.
- Automated Inspections: Computer vision-equipped drones can autonomously inspect equipment, structures, and infrastructure, identifying defects, leaks, or potential safety issues without human presence.
- Predictive Maintenance: predicting equipment failures before they occur by analyzing visual data such as temperature gradients, vibration patterns, or component wear. This helps minimize or prevent accidents or potential environmental contamination resulting from equipment malfunctions.
- Air and Water Quality Monitoring
- Robotic Assistance: Handling toxic substances or operating heavy machinery.
Finally, in other cases, computer vision is not only able to substitute human workforce but can increase performances. This is the case in quality control where a human is bound to get tired and experience a decrease in their ability to detect defects at the end of a workday whereas a computer vision algorithm will sustain the same (and often better) performance.
Across all these use-cases and many more, computer vision is estimated to be a $12Bn-$15Bn marketgrowing at 15% a year, the largest subsectors being Agriculture and Manufacturing.
This is driven by staff shortage as we discussed and the move to re-industrialization of Western Countries. In the past years, countries in Europe and North America all stated a goal to re-industrialize their countries and increase the share of products being produced locally.
As an example, the US manufacturing sector is currently experiencing its most significant upswing in 35 years, and countries like France announced important financing plans (54B€ until 2030) towards re-industrialization.
But as we explained, the biggest limiting factor is the lack of workforce in these countries. It’s difficult to find qualified workforce and costly to recruit and train them.
Headquartered in Ghent, Belgium, Robovision is at the forefront of computer vision: determined to ensure its accessibility across all industries, regardless of sector-specific distinctions. Robovision has crafted a robust yet user-friendly platform, enabling operators without deep technical expertise to create their own deep-learning solutions effortlessly. This platform automates tasks like quality control, robotic object handling, and an array of other automated functions.
But I have to say that my story with Robovision didn’t start well. I set out to meet the team in Ghent, but due to delays, I arrived in Brussel rather late, missed the last train to Ghent, had to take a taxi in the middle of the night only to realize that that key the hotel had left for me was bent and didn’t allow me to get in ! (I managed in the end)
After a short night, I did have a fantastic day with the team the next day. Abel and I went into a lot of details with Jonathan, Thomas, Tim and Florian, who shared their achievements, short term plans and long-term vision.
As usual, we followed up with many calls with experts in Europe and US who confirmed that not only is Robovision addressing nicely the issues we discussed above, but their product has an ideal positioning: an end-to-end, no-code solution for machinery manufacturers and production lines across a multitude of sectors.
It’s tempting for an AI first company to design a product that will appeal primarily to Data Scientists (and that’s the mistake that a few competitors are making from our perspective) But Robovision has understood that a critical success factor for this automation is the empowerment of their end-users such as factory operators and farmers. With Robovision, they are are able to create and maintain their own AI models without having a PhD in Machine learning !
We mentioned labor shortage in Agriculture at the beginning of this article: Robovision has been particularly effective in the agricultural space: ISO Group is using Robovision to plant 1 billion tulips annually (half of the world production).
And they have been a trusted partner to Japanese giant Hitachi in helping them produce semiconductor wafers.
It’s now time to accelerate Robovision’s U.S. expansion so Red River West was a natural partner to foster a robust local presence there and address the increasing demand for automated solutions in American factories.
Our Due Diligence showed the desperate need for this kind of solutions, so we’re very happy to invest in Robovision $42m round, alongside Target Global and Astanor.
As we navigate the complexities of the modern world where staff shortage could hinder the growth of the economy or simply our capacity to feed the planet, we’re embracing the power of computer vision and there won’t be any shortage of people at Red River West to roll up their sleeves and work tirelessly to support Robovision’s expansion to the US.
OH.
