Different automatic systems are becoming increasingly common in land, sea and air transport. Passenger cars are already using driver-support systems, such as Adaptive Cruise Control (ACC). Lane keeping systems are also on the way. Ports are one leap ahead in the development. In a partly closed environment, such tasks as piling containers can be done autonomously using remote-controlled cranes. Many sectors are considering similar opportunities for automating operations. In any automation solution, the technology should closely reflect human activities.
Without paying attention to humans or the environment, automation solutions cannot be made safe and functional. Particularly in transport and logistics-sector companies, ensuring safety is a clear operative goal. However, not every situation can be tested in the real world, so modelling and simulation are also needed. In design, the purpose of research is to ensure in advance a fully functional operational concept and safety with various use scenarios. Sufficient level of safety must also be proven to authorities, and, therefore, the automatic functions must be sufficiently transparent.
What is the sensible degree of automation to aim for?
Autonomous action refers to a system that functions in an independent manner. Still, when talking about autonomous functions, in reality, the goal is not to fully detach the system from human monitoring and control. Even in solutions with a high degree of automation, there are incidents in which maintenance work will probably need to be carried out by humans in the future as well.
Mines are an example of a sufficiently closed environment that can be automated to quite a high extent. The next step in mining, however, would be developing capacity to react to changes without needing to stop the process. In more dynamic and uncertain environments, such as forests, it is more of a challenge to find automation solutions exactly because of the changing and unpredictable circumstances.
The debates about self-driving cars and unmanned ships give rise to such questions whether people will be needed at all and will the number of jobs decrease? In the light of existing examples, automation will replace the jobs disappearing from the working environment with new kind of work, such as tasks requiring coordination capacity and creativity. In other words, humans will still remain a part of the overall picture. Therefore, when developing systems, it is important to decide how people interact with the system and how they receive situation awareness data of automation and the environment surrounding it. How will the user understand how to operate in sudden emergency situations? It is also important that the work remains sufficiently interesting.
Artificial intelligence learns things – but will it learn to become reliable?
Artificial intelligence is supposed to learn more as it is being used. How do we ensure that the learning advances in the right direction? Is it possible to go through a sufficient number of different situations in advance? Verification and validation is one area of expertise which the research focuses on.
It is important to make efforts to build a suitable level of user confidence towards automation. There are many warning examples in history of too much confidence in automation, such as ship wrecks. Therefore, a user should not expect too much from a system, or the kind of capabilities for which the system was not designed in the first place. The other extreme is lack of confidence in automation, which may adversely affect the efficiency of the whole system. Here again, confidence and the human-automation relationship people play a central role in the development work.
In addition, it is extremely important to understand human capabilities and limitations in relation to various user interface solutions. It is important to integrate the human perspective and user-centred methods into research and development of automation technologies intended for professional use and different transport environments. Such methods have also been successfully applied for a long time in the design of consumer products and services, such as mobile phones and websites.
The Research Alliance for Autonomous Systems (RAAS) extensively promotes the automation of transport sector and logistics in particular. In this ecosystem, intended for accelerating innovation within business, in addition to its coordinator role, VTT is also responsible for operational-level design matters. This area covers solutions with differing degrees of automation, such as the development of functional operating concepts, verification and validation of safety, and human-automation interaction. Furthermore, VTT plays a central role in solving ethical questions related to automation, as well as in innovating solutions related to connectivity and safety of cyber-physical systems.