Open technologies will democratize AI

The ongoing digitalization and AI-driven change of the global economy, national economies, and corporations has started and seems to have no end in sight. This change represents societal disruption with many impacts. Continuous change, development and experimentation is the new normal. In order to stay competitive, organizations need continuous exploration of opportunities to exploit data and AI technologies to improve existing business processes and offerings, as well as to find new ones.

A recent PWC report estimates Artificial Intelligence (AI) could make $15.7T potential contribution to the global economy by 2030[i].  The same report identifies nearly 300 use cases for AI spanning business and society.  Finland’s goal to become a leader in applying AI represents an ongoing digitalization and societal change[ii].

AI and the information, communication and automation technologies used in its realization are developing at a breathtaking pace. Development is so fast that education systems face challenges to meet rapidly changing skills needs in the training of workforce to the labor market. Various on-line courses and mini-degrees have increased their popularity in response to rapid skill development needs[iii].

Availability of open AI technologies and related pool of experts has been growing steadily over the past few years.  In 2017, the GitHub community for open source software developers reached 24 million developers working across 25 million repositories of open source code[iv].  Open AI technologies have become a serious option for commercial AI technology offerings.

For example, Google has opened the source code of its machine learning platform behind its own production services, which has created a significant developer and user community around it. In 2017, TensorFlow and TensorFlow Models were two of the top ten most active code repositories on GitHub. Several other AI technologies have also become available with open source licenses.  Just under half of the 100 largest companies in the United States (by revenue) use GitHub Enterprise to build software.  Furthermore, to address the AI skills shortage, globally only 5 thousand teachers and 500 thousand students actively used GitHub in 2017.

Development of new services requires strong AI technology expertise

VTT and IBM Research – Almaden are in research exchange co-operation at Silicon Valley. The goal is to study architecture, ecosystem and future development of open AI technologies from the viewpoint of AI systems development and engineering. Preliminary results of the work are published on a joint blog ( and the topic is discussed at the international OpenTech AI Workshop in Helsinki.

The advantages of open AI technologies include rapid pace of development. The research activities on the field of AI produce new algorithms and machine learning models. For reproducibility of results, these are often implemented and made available with open technologies first. In addition to open source code, lot is happening also around open datasets, machine learning models, benchmarking and leaderboards. The ecosystem around open AI technologies has emerged and is evolving rapidly. This evolution is not only worth following in the sidelines, but calls for active participation to research, development and exploitation of open AI technologies. Clarifying the role and importance of open AI technology for any organization is wise preparation for the future.

The evolution of open AI technologies is a development that has emerged during the past few years. This is continuation to the open source development in software products, which started already earlier. The open development on AI technologies is democratizing opportunities for exploitation of AI; It enables building needed skills, code sharing and exploitation independently of individual vendors in an open ecosystem. Also in the field of AI, value creation and commercial competition are shifting from software products to applications and related services. Here crucial is strong and versatile expertise on AI technologies and capability to apply new and rapidly evolving technology together with customers.

Daniel Pakkala
Principal Scientist, Data Driven Solutions, VTT

Jim Spohrer
Director, Cognitive Open Technologies, IBM

For more information:

[i]  PWC (2017) Artificial Intelligence Study.
[ii] VTT (2018) Finland AI Strategy.
[iii] For example, for a freely available, easy to access online set of courses see
[iv] GitHub (2017) State of the Octoverse.


Robot cars – a future mode of transport

When your car moves without a driver, it suddenly begins to feel like a human and you even start talking to it. Will self-driving vehicles soon become part of everyday life? Will we be hopping into robot cars in the near future?

In May, VTT organised a customer event, where people had a genuine opportunity to find out how it feels to travel in a driverless car. A ‘safety person’ was present during the demo situation, but his only task was to monitor how the robot car behaves and enjoy amazing engineering work. However, humans are still needed to decide on the functionality of the vehicle but it has been pre-programmed as kind of a law book telling the car how to act in each situation. Time has passed since the spring day event; in the intervening period more laws have been inscribed in the book programmed into the cars, Marilyn & Martti. The next milestone was reached on 27th of June, when the silence between the engaged couple ended and the cars began exchanging data on their relative positions and avoiding each other with using the ITS G5 communication channel. Kisses are strictly forbidden, since each one would cost more than EUR 15,000.

Matti Kutila VTT

The robot cars Marilyn & Martti have provided tangible proof that motoring for future generations could and will be totally different to what we are used to. Young people use information technology, tablets, mobile phones and social media differently to those of us whose first computer was the legendary Commodore 64, acquired mainly for playing games. Cars may be here to stay, but they will be used differently – they are an efficient mode of transport that will be integrated into the transport system by the IT skills of young people, their adventurous spirit, and the availability of the ICT ecosystems. Automation will also allow drivers to become passengers which people are ordering using their mobile apps and IT clouds.

Matti Kutila VTT

Full automation is still some way off

When driving my first car (a Skoda 120 LS) in 1992, I could not imagine that, just 20 years later, my car might automatically park itself as my current car (a Volkswagen Touran) does. Similarly, in 20–30 years’ time my sons will reminisce about how their dad even drove his car in a city bustling with pedestrians, and had to do his own braking at red lights – making them to wonder how inefficient and unsafe was that era!

When I turned 18, it was obvious that I needed a driver’s licence and a car of my own. My sons may not think in the same way, since for them renting or sharing a car will also be an option, as long as they can get around – mobility will trump ownership. They are likely to obtain a driving licence of some kind, but their driving school will probably be different from mine, where the biggest challenges were memorising road signs and remembering to give way at hospital intersection in small city called Forssa. When switching on their car’s automated mode, my sons will probably need to learn how to supervise safety of the automation system and handle status messages of the in-vehicle computer units.

Matti Kutila VTT

Although I strongly believe that the future of road transport lies in automation, the transport system is unlikely to change overnight. This will occur through normal technical evolution, in which automation is introduced step-by-step one aspect, function and area at a time. Full, 24/7 automation everywhere, in any weather conditions and available for all is still 20–30 years ahead, but that is where development is taking us – I have no doubt about that. It is great to be part of the group that is moving this revolution forward and especially, be in the driver’s seat with my world-class colleagues, the VTT RobotCar Crew.

Matti Kutila_Citroen_01072016

Matti Kutila
Automated Vehicles, Research Team Leader
Twitter: @matti_kutila

Robotic cars are coming – are we ready?

Matti Kutila_Citroen_01072016

Robotic cars and automation in traffic were the great innovation of 2016 – or were they really? Could they perhaps just represent a long-term revolutionary path that has merely attracted special attention in recent media headlines? In any case, automotive engineering is undergoing transition from metal bending towards software development which is extremely fascinating.

Cars travelling 24/7 in a snowfall on icy roads are a future dream and more reliable technology is needed to equal surpass human driving skills. For example, a self-driving car representing the current state of the art cannot enter a multi-lane roundabout in congested Paris traffic – or at least leaving roundabout would be enormous challenge.

Automation in traffic is now top of the famous Gartner hype curve which is a measure of technology interests in industry. Unfortunately, the next few years a steady downward slope is expected before a slow rise again. Meanwhile, the automated functions will take over control of the vehicle step by step and becomes a master of driving instead of being assistants.

Development of robotic cars began more than 35 years ago

Although the media was awakened to the significant investments made in robotic cars by Google five years ago, the seeds of automation had been sown considerably earlier. Mercedes Benz introduced the first self-driving car in 1980, using the technology of the time. However, the loudest starting shot was fired at the DARPA Grand Challenge competitions, arranged by Pentagon in the U.S. between 2004 and 2007. The core of Google’s development teams was also made up of the university teams which had successfully participated in these completions.

Starting from the early 1990s, the European car industry has brought to the market a series of active, electronics-based safety systems. Through EU projects, we have had ringside seats to follow the development. Automated safety features now being released onto the market can be considered as the next generation of active safety features. Thus, the development of automation did not begin five years ago – it was initiated already 35 years ago.

Difficult road weather conditions pose a problem

Despite the product development efforts worth several billions undertaken by the automotive industry, traffic authorities and public funders, the world is not yet ready. As a matter of fact, at the current stage, fully automated vehicles are fairly primitive. Such vehicles are capable of travelling on roads at up to 50 kilometres per hour in areas covered by an accurate mapping data and during sunny weather. Cars travelling 24/7 in a snowfall on icy Finnish roads are still a distant dream, perhaps ten years away.

The current sources of sensory data are not sufficiently reliable in harsh weather conditions, and the processing capacity of vehicles to understand varying traffic incidences is far from the human brain. Not until now has the technology reached a point where sensors for talking and hearing can be installed, thanks to the capability of vehicles systems which enables them to exchange data. An obvious demand for automation in traffic exists, but the technology still needs further development. Anyway, having an opportunity to develop cars of which the public only has a faint conception is a researcher’s dream job.

Mixed traffic poses problems

The widespread adoption of automated cars is often said to improve traffic safety – after all, automation removes one of the factors underlying accidents, the human error. Traffic is also expected to run more smoothly as automated vehicles can travel closer to each other than cars with human drivers. Automated cars also facilitate the travel of such people who cannot or do not want to drive a car for some reason.

As we have had the opportunity to read during the past few weeks, gains in safety can be made, but not even automation is able to deal with every possible situation. One unfortunate collision involving a fatality has already occurred, as the sensory system of a car controlled by an autopilot failed to recognise an obstacle ahead of it. One additional challenge to safety is the fact that robotic vehicles do not travel in traffic composed of their likes, but in mixed traffic involving ordinary cars, pedestrians and cyclists.

Making traffic run smoothly requires not only automated systems but also a capability of vehicles to ‘talk to’ each other. An automated vehicle just following the car in front of it and reacting on its movements is not enough. After all, a good and experienced driver follows the traffic farther ahead, and in this way is capable of anticipating new situations. At present, human drivers are more flexible than automated cars.

What next?

The technical development of automated cars will continue and the price of its components will come down. More and varied experiments will be conducted, providing excellent data for impact assessment. At the present stage, impacts can only be assumed, with scenarios representing the opposite ends of the spectrum being equally likely. Therefore, it is important to be involved in development and research.

Components for smart cars have been developed for 30 years, and will be developed for the next 30 years. The decades to come will present challenges. The price of the components representing the previous generation will decrease, with new features being introduced in the high-end cars. Alternatively, we might be mistaken, and the traditional car industry loses the game and the new car brands will bear names such as Baidu, Google, Apple, and Tesla. The future manufacturers might even deliver a set of components to the customer who could then assemble a car by following instructions, in the same way as the furniture industry delivers its products today. Of course the automotive software is being sold separately for a monthly fee. Let’s wait and see.

The only certain thing is that traffic philosophy will change. We already have mixed traffic involving traditional cars, automated vehicles, more or less automated lightweight vehicles (such as Segway PT electric vehicles), and public transport. This will make the traffic environment highly complex. In order to help researchers to understand the changes in guiding product development in the right direction, various field tests involving motorways, urban environment, networked environment, closed areas, crossroads, winter conditions, articulated platoon of lorries and passenger cars need to be conducted.

Building and maintaining traffic infrastructure and the R&D of vehicles is expensive and requires a great deal of work. Field tests will substantially reduce the risk of wrong or unnecessary investments. Tests not only offer an opportunity to assess impacts and to provide support to companies conducting R&D, they also provide an excellent channel to communicate to the public the true meaning of automated traffic and to prevent people from getting the wrong ideas and attitudes. Automated vehicles are not developed for engineers or authorities, but for ordinary people in order to enable them to travel in a more convenient way in future.

Merja Penttinen

Matti Kutila, Senior Scientist 

Merja Penttinen, Research Team Leader; Twitter: @MerjaPenttinen