IFFI

Knowing where to start with AI

Artificial Intelligence has proven its value, with most companies now recognizing its importance. However, the challenge lies in how to implement AI within your own company and where to start. Ben Lomax Thorpe and Sjors van Gool from Royal HaskoningDHV delve into the AI landscape, emphasizing its significance in the engineering domain and its impact on both RHDHV and its clients.

Several reasons can cause companies to be hesitant about AI implementation. Lack of knowledge or the right personnel, as well as fear or uncertainty about AI, are common barriers. Questions about privacy also arise. “It’s essential to start with the challenges within your company to improve, not the wish to apply AI,” says Van Gool. “For instance, streamlining processes, making better predictions regarding purchases or sales or improving operational processes within the company. Determine where the added value lies for your company and evaluate if AI can realise this value. Additionally, critically evaluating key factors that make reliable application of AI possible is important, for example data governance within your company is crucial. Since a lot of AI algorithms rely on collected data, the quality of predictions depends on the quality of available data. Ensuring reliable data is, therefore, an important initial step.”

One of the applications aiding in predictions and process optimization is Digital Twins. “These create a digital replica of an asset, process, or system, assisting in better understanding and optimizing its performance. For example, a digital twin can optimize the performance of a robot, production line, or even end-to-end supply chain,” explains Lomax Thorpe.

A digital twin has two purposes: 1. to help us understand the physical world by providing accurate insights into the status of the entity or process it represents; 2. to support informed decision-making that helps us adapt and change that physical reality to make it better. “To deliver these capabilities, digital twins consist of various building blocks. These building blocks, and their underlying technologies, need to be implemented as part of your digital transformation, combining people, processes, and technology,” says Lomax Thorpe.

The digital twin is fuelled by real-life data, offering an accurate depiction of the current situation and future projections. “You can leverage the digital twin to simulate various scenarios, showcasing how and where processes can be adjusted within the factory to achieve desired outcomes.” The advantage lies in experimentation and adjusting various parameters before making actual changes in the factory. Successful implementation of a digital twin requires human assistance. “Real-world knowledge is often necessary to aid in decision making; a digital simulation is not sufficient.”

For those wary of technology, this might be reassuring. AI and digitalization can greatly benefit your company, but human involvement remains (for now) essential. Do you want to know more about digital twins? Do you want to know how to effectively start with AI within your company? And are you curious about what Royal HaskoningDHV can offer? Join us on April 25th at the IFFI event in Zwolle.

Sjors van Gool

Ben Lomax Thorpe