The term “Smart Lab of the Future” describes a fundamental transformation in research and development. Series of experiments and handling tasks on various devices run fully automatically and autonomously in the lab. An AI agent assists in planning experimental procedures. Multiple lab sites are connected via self-driving vehicles, enabling flexible sample exchange. This is what the future might look like in a smart development lab. Automated processes, artificial intelligence (AI), and networked lab infrastructure are profoundly transforming research and development processes. Smart Labs stand for robust, repeatable processes, transparent data, and accelerated development cycles – key levers for lowering barriers to transfer between the lab and production. For companies, this means lower scaling risks, faster validation of hypotheses, and seamless integration into existing quality and compliance systems.

The “RESEARCH TO BUSINESS live” event on September 16, 2026, took up this trend and brought it to life for approximately 30 guests from industry and commerce right on site, in the laboratories of various KIT research groups. At five compact stations, researchers provided insights into autonomous laboratory platforms, AI-supported control systems, digital data flows, and much more. The breadth of the research presented demonstrated just how diversely the concept of the Smart Lab is being explored at KIT. The lab tours showcased specific application scenarios, while at the same time emphasizing direct dialogue: What technological building blocks already exist? What are the current hurdles? And how can companies get involved in the development at an early stage and integrate it into their own research and development? The event, attended by guests from the fields of computer science, automation, laboratory infrastructure, and logistics, underscored the relevance of the topic for a wide variety of industries.

Smart Lab 1: Laboratory Platform for New Semiconductors

The event kicked off at the Center for Materials Science and Engineering for Energy Systems (MZE) at KIT. Prof. Dr.-Ing. Alexander Colsmann and Dr. Holger Röhm from the Institute of Lighting Technology (LTI) presented E-MAP, an autonomous laboratory platform for researching energy materials. Coating, sample handling, and analysis are integrated within this platform. A key focus of research at the LTI is on more sustainable manufacturing processes. Röhm reports: “We are investigating how organic nanoparticles can be deposited from water or ethanol in order to avoid the use of toxic solvents as much as possible. These processes enable the environmentally friendly production of very thin layers and offer prospects for large-scale production.” In this way, the platform addresses both the ecological and economic requirements of modern materials development.

E-MAP demonstrated how a largely autonomous environment can significantly accelerate material development – an aspect of great relevance to the semiconductor and electronics industries. The researchers emphasized that automation and machine learning should be deployed strategically where bottlenecks exist – such as in repetitive sample handling and quality control. The role of intelligent automation in sample handling was particularly evident. Cameras monitor the coating process, a neural network detects the position of the samples, and a robot receives grasping instructions derived from this data. The goal is a robust, reproducible process that reduces errors while conserving resources. For companies, this approach opens up opportunities to examine development processes more systematically and to test a larger number of variants under reproducible conditions.

Smart Lab 2: Mobile Robots Connect Laboratories

The second stop at the Institute for Material Handling and Logistics (IFL) demonstrated how material samples will, in a sense, transport themselves in the future. Using “Cat001” and “Mouse001,” Lukas Müller and Hao Pang demonstrated how autonomous vehicles can transport samples between different locations and organize them flexibly. The researchers’ broader vision is a networked infrastructure in which many decentralized “self-driving labs” are connected via a shared cloud. This would allow different research sites and laboratory processes to collaborate without requiring all tasks to take place at a single central location. Müller emphasized: “Logistics is not merely a supporting task; it can be a key prerequisite for the networking of autonomously operating laboratories.”

The technology presented uses laser scanners to continuously map the environment. The robot-like vehicles create maps of their surroundings and determine their position in real time using algorithms. A photorealistic digital environment is designed to assist with navigation. This complex digital infrastructure raises further questions. Topics such as secure data rooms, digital sovereignty, and the collaboration between humans and AI in complex research processes were discussed with the guests. The “human-in-the-loop” principle combines the capabilities of autonomous systems with human control and responsibility. Here, humans remain involved at key points.

Smart Lab 3: From Material Discovery to Application

A visit with Prof. Dr. Ute Schepers, Institute for Functional Interfaces (IFG), highlighted the potential contributions autonomous laboratories could make in the future to medicine and health research. In the Bio.MAT project presented, autonomous laboratories are used for the exploratory search for new biomaterials. Particular attention was paid to printable bio-inks – a material that plays a central role in the three-dimensional printing of biological structures. Automated production and testing of material combinations, along with the parallel analysis of large sample sets, can help systematically investigate new biological materials. The research approaches presented ranged from heart valves for babies that grow with the child to sensors – known as “Earables” (Wearable Computing on the Ears) – that can monitor health parameters. These examples illustrated how automated platforms can support research into complex materials and shorten development cycles.

At the same time, an issue emerged that directly affects many industrial companies: “The technical hurdle often no longer lies with the software and AI itself, but rather with the lack of interfaces between devices, software, and data platforms,” said Schepers. Many devices have not yet featured open interfaces for AI software, which complicates automated control. For the next generation of autonomous laboratories, license-free, standardized device platforms and open integration options are becoming increasingly important. This presents a clear area for innovation for suppliers of laboratory hardware: “AI-ready devices” can become a central linchpin of autonomous laboratories.

Smart Lab 4: Software Architecture for Self-Driving Labs

Prof. Dr. Matthias Franzreb demonstrated at the IFG what such digital laboratory ecosystems with AI and integrated devices might look like. He highlighted the software architecture and, with it, the ability not only to automate experiments but also to plan and optimize them entirely. In a live demonstration, an experimental protocol was formulated in a manner similar to an input for a dialogue-oriented AI system. The AI then developed a sequence for the laboratory activities and also took over control in an experimental setup. These agent-based systems can, for example, plan experiments, create necessary device scripts, conduct experiments, evaluate results, and use them to design the next experiment. According to Franzreb, this shifts the role of AI: “It serves not only to analyze existing data but can also be integrated into research workflows themselves. This makes AI an interface between scientific questions, laboratory software, and experimental practice.”

According to Franzreb, a key prerequisite for practical application is a clear separation between AI-supported planning and controlled experiment execution. The actual execution of the experiments remains deterministic and is not left unchecked to the AI. The AI supports planning and optimization, while execution follows defined procedures. This builds confidence in reproducible processes and, at the same time, enables the use of AI’s potential to increase efficiency. For user interaction, the presented software architecture relies, among other things, on wizards and low-code approaches. This is intended to allow even those without extensive programming knowledge to work with the system and integrate additional devices more easily. The combination of intelligent planning, user-friendliness, and device integration is a particularly relevant aspect for companies looking to further develop their existing laboratory infrastructures.

Smart Lab 5: Data as the Foundation of the Autonomous Laboratory

The final stop in the Self-driving Lab ChemASAP at the Institute for Biological and Chemical Systems (IBCS) brought together many of the aspects presented earlier. In the organic synthesis lab – which will be fully automated in the future – Prof. Dr. Stefan Bräse and Dr. Patrick Hodapp demonstrated how the entire process, from reaction preparation through synthesis to analysis, can be digitally supported and documented. Materials are developed and characterized at high throughput without manual intervention. Parts of the room-filling laboratory were presented, including a transport system with magnetically guided transport paths that ensure highly precise positioning of samples during the experimental process. Dual-track operation without fixed routes enables flexible material logistics without congestion issues. The solutions presented made it clear that autonomous laboratories are not solely based on robotics, but rather on a combination of data management, software, and automation.

The starting point for the development of ChemASAP was the question of how experiments could be documented in a way that ensures long-term traceability and reproducibility. Traditional electronic or handwritten lab journals were often insufficient, explains Bräse: “The digital lab journal Chemotion, which was developed by Dr. Hodapp and Dr. Nicole Jung and is used for ChemASAP, therefore records not only experimental data but also equipment configurations, lab setups, installation parameters, and process data. The resulting digital knowledge base forms the foundation for automated research processes and future reproducibility.” Building on this, automated systems for synthesis, analysis, and materials logistics were developed step by step, offering seamless data management. The benefits for research and industry are clear: the better data is structured, recorded, and made available over the long term, the easier it becomes to ensure reproducibility, preserve knowledge, and advance technological development. The result is also an extensive molecular archive containing a collection of 18,000 substances – including long-term archiving of samples and verification of chemical identity and stability.

Smart Labs as a Shared Challenge

Despite their different areas of focus, the projects were united by a common question: How can research and development be accelerated through intelligent, data-driven, and automated processes? Many of the examples presented demonstrated that autonomous laboratory processes are already possible today; at the same time, however, standards, safety, and interfaces set the framework for broad industrial implementation. Their success depends largely on close collaboration between research and industry.

In addition to technological insights, direct exchange between research and industry was a central component of the event. The five stations showcased different stages of development, research questions, and fields of application related to autonomous laboratories. It is precisely this diversity that sparks innovation: For the participating companies, the event offered the opportunity to learn about current developments at an early stage, discuss challenges with researchers, and identify potential starting points for joint projects. In this way, RESEARCH TO BUSINESS live – Smart Labs of the Future exemplified how technology transfer thrives on direct exchange and shared visions for the future.

The RESEARCH TO BUSINESS team would like to thank all guests and the participating researchers who opened their lab doors to us. Throughout the day, a sense of curiosity about the new and an openness to direct exchange were palpable.
 

Anyone interested in learning more about KIT’s technology transfer activities can find information about the offerings of RESEARCH TO BUSINESS online. Current activities and projects can be found on the blog. The newsletter and the Alert also provide information on new developments and patented technologies from KIT for which we are specifically seeking industry partners.