AI and autonomous systems improve safety, efficiency and sustainability


If machines, such as industrial systems, cars, buses, trains, ships, etc. can be automated to provide increased safety, efficiency and sustainability, why wouldn’t they? Autonomous systems may be the next game-changing megatrend where Finland can fully utilise its world-class ICT, machine and system building expertise.

The world needs compelling visions, solutions that can improve safety, environment and quality of life. In many fields, such as computer vision and natural language processing, AI and machine learning methods already enable solutions far beyond conventional industrial automation.

The change will come inevitably little by little anyway. Supporting this trend is important since steering is always easier when one is sitting on the driver’s seat.

Autonomy will be reached gradually

As the lifecycle of these equipment is long, there is a long transition period from current systems to fully autonomous systems. In cases where transitioning is done from manned to unmanned operations there is enough time to transition people to new skills in the next tens of years.

Autonomy will not be reached in sudden leaps, but rather by gradually developing systems towards full autonomy. Also, not all such systems need to perform in the same way as human operators.

It is likely that as for example vehicle autonomy increases, communication and collaboration between autonomous agents can be used in ways not possible with human operation. Fully autonomous systems (within a described mission) may represent significant improvement in safety and efficiency – and ultimately in entire business models.

Collaboration between humans and partially autonomous systems is useful

The good news is that even currently available off-the-shelf AI solutions are helpful and enable collaboration between humans and partially autonomous systems.

Self-driving cars are not here yet, but adaptive cruise control and lane assistant are. Autonomous ships do not navigate the seaways yet, but there are commercially available means of enhanced situational awareness that may significantly improve navigational safety.

As data is collected by such assistant systems in real environments, new models can be built using for example artificial neural networks to provide automation systems with increasingly detailed understanding of natural environments.

New types of data is created

Autonomous machines and their operating infrastructure will create new types of data that are not yet captured by the few currently dominating data platform players. This is likely to be important to acknowledge, if we want to take the driver´s seat in the development. The emergence of 5G communication and autonomous traffic even at low autonomy levels generates vast amounts of data that does not exist today in any data platform.

It is important to understand early on how such data and related metadata need to be collected, organized, and annotated in order to facilitate long-term development of autonomous systems. Significant practical necessities for data collection include machine-readable formatting and synchronization of all collected data, and human annotation of data for which accurate machine learning models are not yet available.

Finland needs to execute compelling visions of autonomous systems

Finland is known for its ICT capability, collaborative approach and the informal relationship between private and public sector. This may prove beneficial if other players represent the old world´s single player zero-sum approach.

Finland needs to execute compelling visions of complete autonomous systems and infrastructure – together and first in the world. The enemy to this is the “something for everybody” approach. Small projects in a small country are unlikely to lead to global scalability.

Finnish AI experts generate global value

Understanding AI big time, each one of us needs to step up and train ourselves in this new skill. We need to demand management, officials, owners and operators to apply available off-the-shelf solutions in our daily business. We should not only solve issues in our small country but aim at significant global value creation, utilizing our natural strengths.

Every Finnish world-class AI expert will generate global value and bring money home. Our education system needs to ramp up to new requirements and public funding to support globally scalable initiatives.

Our most successful moonshot project may be the one that enables others to go to the moon – our earning logic is to derive income from others’ value.


The writers belong to the ecosystems working group of the AI Finland Programme.

About the authors

Merenluoto Jukka

Jukka Merenluoto

Ecosystem Lead at DIMECC Oy

All posts

Poikonen Jussi

Jussi Poikonen

AI Development Lead at Rolls-Royce

All posts

Eloranta Sauli

Sauli Eloranta

Head of Innovation & Technology at Rolls-Royce Marine

All posts

What do you think?

Mitä mieltä sinä olet?