Robovision’s team (left to right): Florian Hendricks (CGO), Jonathan Berte (Co-Founder and Chairman), Tim Waegeman (Co0-Founder and CTO) and Thomas Van den Driessche (CEO).

The Future of Industrial Automation and Our Investment in Robovision

Target Global


By Bao-Y Van Cong & Tobias Winczer

In the rapidly evolving landscape of industrial automation, the convergence of computer vision technology and artificial intelligence (AI) stands as a transformative force. With surging production costs due to higher labor costs, energy prices and more complex supply chains, increasing competition due to globalization, and a severe lack of specialized industrial workers, companies are continuously looking for new ways to harness the power of automation and AI in order to boost production efficiency and automate manual operations. Recognising the breadth and depth of use cases that AI offers across the industrial space, we delved into the relationship between computer vision and industrial automation and its potential to elevate the competitiveness of established economies, unlocking unprecedented innovation across a multitude of sectors. Here is what we found.

Unlocking AI-powered Industrial Automation at Scale: This is Robovision

Imagine a world where the expertise of seasoned assembly-line operators seamlessly transfers to machines through innovative Vision AI algorithms, streamlining manual tasks. Now, imagine AI engineers dedicating their time to developing new algorithms rather than wasting countless hours with routine implementation and maintenance of models in production.

This vision seemed far-fetched until Robovision came along. On a mission to democratize and make Vision AI accessible to all stakeholders in a truly industry-agnostic way, Robovision built a powerful but easy-to-use platform that empowers operators with limited technical expertise to develop their own deep-learning-based solutions, automating tasks such as quality control, robotic object manipulation, and a wide spectrum of other automated actions.

Robovision AI-powered smart planting machines from ISO handle half of the 2 billion tulip bulbs planted annually in the Netherlands.

Serving customers across a multitude of industries — from foodtech, medicine, oil and gas to construction and precision agriculture — Robovision’s platform enables applications such as 3D surgery planning, construction materials sampling, laboratory plant propagation, and real-time monitoring on drilling sites.

A rocky road ahead: The transition from old-school computer vision to AI-driven automation

In the realm of high-volume production facilities, static computer vision technology — think cameras for production line monitoring — has long been the go-to solution for automating processes. However, as businesses continue to strive for more actionable insights and dynamic automation, the spotlight is now on vision AI to elevate these conventional use cases and make them more “intelligent”. Yet, despite its potential to revolutionize mission-critical processes across various industry verticals, the adoption of Vision AI comes with its fair share of hurdles.

One significant challenge lies in the transition from the experimental to production phases, with a mere 10% of machine learning initiatives making the leap. This is due to the fact that many companies face limited IT budgets and lack sufficient AI talent, causing a disconnect between R&D labs and field operations. Traditional businesses, wary of the significant uncertainties and prolonged timelines associated with AI projects, often hesitate to invest, leaving the realm of AI-driven automation largely monopolised by a handful of tech-first companies with sufficiently large and robust AI engineering and data science teams.

Resorting to off-the-shelf software solutions doesn’t cut it either. Most existing software vendors offer generic one-size-fits-all products tailored to serve narrow vertical use cases and are ill-suited for the complexity of enterprise environments. To truly unlock the potential of vision AI means enabling customers to retain autonomy over solution development and provide sufficient configurability for their specific use cases. On top of that, enabling vision AI in Industry 4.0 often poses requirements such as flexible application deployment on edge devices, integration with existing hardware, real-time data streaming for use cases with strict latency requirements, or advanced 3D vision capabilities. As a result, the bar for implementing effective vision AI is restrictively high, leaving industrial businesses reliant on outdated technology and manual labour.

Meeting the rising demand to build tailored, proprietary vision AI

Instead of using generic 3rd-party software, industrial businesses have a strong appetite to build in-house proprietary algorithms powered by their own data and embed these into existing workflows in order to gain a long-term edge over competition. However, in attempting to do so, they tend to run into the following challenges:

  1. Lack of technical resources in AI development — While many enterprises might have in-house data science teams capable of experimenting with machine learning on a proof-of-concept level, these teams rarely have the capacity to build full-scale production-ready ML applications. This creates a pressing need for a platform that streamlines the end-to-end model lifecycle, allowing customers to turn AI experiments into production-ready applications with real ROI.
  2. Lack of autonomy in production — The true value of AI projects unfolds when models in production are continuously retrained and improved. Whenever the use case or input data changes, new training data must be acquired and labeledlabelled, models must be re-trained, re-evaluated, and re-deployed. In practice, most customers have complex setups, with machines requiring daily retraining, with 10–100k predictions as input. However, as non-technical machine operators can’t handle this, they depend on data scientists for maintenance. This presents an opportunity for a low-code environment to provide non-technical end-users with the operational autonomy to maintain and operate models in production.
  3. Integration with hardware and the ability to run inference in real-time — Industrial applications often require full integration of vision AI into existing hardware systems (such as control centres and monitoring cameras) and need to produce inference with low latency, even in challenging low-connectivity environments such as oil & gas rigs or industrial greenhouses. As a result, Vision AI software providers for the industrial space need a robust understanding of not only software but also hardware and edge computing.

Blending its domain-specific capabilities with cutting-edge AI, Robovision addresses these challenges head-on — enabling industrial businesses to embark on a journey towards enhanced efficiency and competitiveness through AI-driven automation.

Revolutionising Industrial AI: Robovision’s robust but easy-to-use ML platform

Robovision’s product is an end-to-end platform enabling customers to easily create and deploy AI models for computer vision. The platform enables enterprises to enhance their standard (static) computer vision setups with tailored vision AI models trained on their own data. Using the platform, customers can easily label their visual data using Robovision’s AI-assisted image annotation tool, set up and train custom AI models, evaluate model performance, deploy these to machines on the edge, and continuously manage, amend and retrain them once in production.

Instead of requiring a large data science team and a complex stack of ML tools to complete this process, Robovision enables customers to handle the entire model lifecycle through an intuitive low-code visual user interface that can be operated by a small team of non-technical machine operators. By orchestrating the entire model lifecycle and decreasing the dependency on data scientists, Robovision accelerates the time-to-ROI of AI projects, limits the upfront resources required and enables customers to operate fleets of AI-driven industrial robots at scale.

Figure 1: Robovision streamlines solution development and enables faster time-to-ROI

Ambition to power AI-driven Industrial Automation Globally

Our journey with Jonathan, Tim and Thomas has been nothing short of inspiring. When we first met the team, we saw an exceptionally strong founder-market fit built over the course of over 15 years in the computer vision space. Having spun out of an established AI consulting business, the team have cultivated deep expertise in building verticalized AI vision systems for specialised high-tech customers.

Plant sorting machine by ISO Group, powered by Robovision

Robovision’s strong commercial traction and customer love are a testament to what Jonathan and the team have built. With their software now powering over 1,000 machines for clients across 45+ countries and generating over $250m in sales for industry giants such as Hitachi and ISO Group, it is clear that Robovision is on the right path to revolutionize industrial automation on a global scale.

We are thrilled to partner with the Robovision team by leading their $42m Series A fundraise, in our first Belgian investment, alongside Astanor Ventures and with participation from Red River West. We look forward to being a part of Robovision’s mission to become the leading vision AI company globally, helping them to drive the fifth wave of the industrial revolution powered by AI.



Target Global

Leading European tech VC with €3B+ AUM. Known for backing fast-growing startups & capitalizing on overlooked opportunities. 15+ unicorns, 21 exits, 7 IPOs.