You are currently viewing Fizyr: Automating Human Tasks by Deploying High-Quality Computer Vision Software 
Fizyr B V | Herbert ten Have

Fizyr: Automating Human Tasks by Deploying High-Quality Computer Vision Software 

There were the days when industrialization was a synonym of textile industry only, and the artisans were the industrial workers. But the industrial revolution invented power looms, and the juggernaut of automation began its journey. Back-to-back industrial revolutions transformed the medieval world into the modern age. The first three of them, i.e., the steam engine, age of science, and mass production, accelerated the world-saving human efforts, costs, and money.

However, the juggernaut of automation kept traveling, and today, it has reached the station, where the globe is digital – the fourth industrial revolution. Human efforts were first replaced by animals, animals replaced by machines, and to control these machines, robots arrived. Today, these robots are deployed in every sector, especially to perform repetitive tasks. These robots can move in six degrees of freedom and operate in X, Y, and Z planes.

In this fast-paced world, time is money, and these robots help to save time by performing repetitive tasks. However, they follow the path which is inscribed by the software programs. Looking at the growing market, many companies started to provide software to automate robotic movements. Fizyr is one such company that offers a standardized software product, based on artificial intelligence, that allows robots to cope with high variation of objects in automated picking and placing applications.

In our quest for ‘The 10 Most Disruptive Automation Companies to Watch’, Mr. Herbert Ten Have, CEO of Fizyr, shared his endeavor with us. He elaborated on the company’s work culture and the future aspects. Following are the details of the interview:

 Please brief our audience about Fizyr, its USPs, and how it is currently positioned as a leading provider of automation solutions.

Fizyr designs, builds, and installs a standardized computer-vision software that detects unknown objects varying in shape, size, colour, material or stacking. Its robust algorithms provide over 100 applicable grasp poses each second, including classification to handle objects differently, while performing quality controls. Just as the camera and gripper are the eyes and arms of a robot, the vision software is the brain that analyzes the image and directs the robot to move accordingly.

Fizyr is trusted as a strategic partner by global leading integrators, thanks to its fully scalable and proven software-only product, for which we provide full insight and control to install, manage, update and support. We empower system integrators to build cutting-edge solutions for the industry’s end users, in multiple high-demand industries, such as e-commerce, micro-fulfilment, parcel handling and more.

 Tell us more about the products and services through which you have ascertained your stronghold within the automation space.

In such a rapidly growing market, logistics companies expect robots to perform their tasks as predicted and agreed. Applying Fizyr’s product is a strategic choice for which we provide full insight and control to install, manage, update and support. We license a standardized product for automated picking, placing, and stacking to system integrators and end users. Our computer-vision software is trained with a unique dataset with millions of images of challenging logistics environments, proposing the best possible grasp poses. Applied in production all over the world, it empowers the robotic automation of leading system integrators and end users, in online retail, warehousing and parcel services.

Fizyr’s software is also hardware agnostic, giving clients the freedom to choose the best camera, robot, and gripper for their automated picking cell. We believe in the value of Fizyr’s independency and want our clients to be able to use open standards and choose their hardware to assemble the best possible picking cells.

Although the company purely focuses on software, Fizyr also licenses the designs to modify, 3D print, and assemble its proprietary gripper. This gripper can handle challenging situations, ensuring ease in integration with any robot in a picking cell.

 Being an experienced leader, share your opinion on why the adoption of technologies such as AI and machine learning is necessary for the automation services industry and how you have leveraged these technologies to deliver your products/services/solutions?

The items and parcels tend to have a huge variation in terms of sizes, materials, colors, and placement, which traditional logistics systems cannot account for. In e-commerce, there are millions of SKUs (Stock Keeping Units), thousands of which cannot be programmed. Therefore, the adoption of technologies such as AI and deep learning has the potential to change the face of the industry.

Fizyr’s neural network is trained to classify an object, strategize the pressure required to pick it, and propose the best possible grasp poses to singulate it from bulk. One of the initial projects managed by Fizyr in 2017 involved enabling robots to work 24/7 and fold up to 720 towels per hour with Laundry Robotics. Fizyr’s computer vision algorithms were able to propose over 100 grasp poses each second on unknown objects.

 Considering the COVID-19 pandemic, what challenges did you face, and how are you currently driving Fizyr to sustain operations while ensuring the safety of your employees?

One of the major challenges faced by the company was the termination of a project by an Italy-based client due to the dire consequences at the start of the pandemic. Consequently, other clients hesitated to continue project operations as they required hardware components, such as cameras, grippers, robots, etc., to conduct research and development. The first few months were uncertain, and later, several clients saw the opportunity for growth in in e-commerce, micro-fulfilment and parcel handling.

Those clients motivated Fizyr to scale up its processes even faster and offer robotic solutions to the dynamic industry. Meanwhile, the Fizyr team was working from home and would only work from the office if absolutely necessary. At that time, the huge office space in the pandemic was reserved for specific purposes. With the restrictions easing down in the Netherlands and the high vaccination rate, people are moving back to the office on a flexible basis.

 What would be your advice to budding entrepreneurs who aspire to venture into the automation solutions niche?

Herbert highlights the importance of validating assumptions, pushing the product to the market, and creating value for the clients in the quickest way possible. The quicker the entrepreneurs identify opportunities, the better they get to approach the clients in an early phase and acquire the market before facing any consequences later. Understanding the requirements of the clients from the beginning can stimulate company growth, improve partnerships, and help build relations with potential clients.

 How do you envision scaling Fizyr’s operations and offerings in 2021?

The fast growth of e-commerce and parcel handling has exacerbated the human labor shortage, which is expected to keep rising. As a response, an increasing number of companies have started to adopt robotic automation and implement solutions for automating picking and placing applications both in e-fulfilment and parcel handling. We believe we have solved one of the biggest challenges in logistics and we hope to empower as many integrators and end users worldwide as possible, in an effort to relieve the pressures of the industry. The product and market are ready, so we plan to scale exponentially in the upcoming years, reflecting such a rapid growth.

Fizyr remains dedicated to training, upskilling, and helping employees thrive in the fast-moving industry of AI and logistics. The company aims to focus on predictable and scalable processes to resolve challenging situations and contribute to the industry.

About the Leader

Fizyr’s CEO, Herbert Ten Have, joined Fizyr (previously called Delft Robotics) in January 2017. Delft Robotics started as a robotics integrator. Within a month of Herbert’s arrival, the company won the prestigious Amazon Picking Challenge. This led to the company pivoting towards a deep-learning computer vision software-only product and re-branding to Fizyr. Herbert is an innovative business leader with more than three decades of experience in building high-tech products and services.

Thanks to his expertise and knowledge of the industry, he geared the brilliant Delft Robotics team towards success, thus, leading to an unprecedented impact on the automation industry. He proactively engaged the company in recognizing industry problems and solving them while staying committed to building the best vision software product and provide cutting-edge robotic picking solutions for the logistics industry.