AI Trends in Mobility Sector

19 May, 2023 15:16 EEST
Matti Kutila,
Ari Aalto

Digitalisation and sustainability are the two mega-trends in mobility domain. Green transport endorses use of alternative fuels and electric vehicles. The automotive and machinery industries has recently launched even heavy good vehicles to urban environment where power train is electric. The second major mega-trend is digitalisation of mobility in all levels starting from single vehicle and end up to fleet management. The operation that traditionally has been done manually are now not only recorded digitally but also processed to suitable format automatically. Recent examples (ChatGPT [2], fotor [3], etc,) have been entering to the markets and they have shown exceptional capabilities to use the existing data in internet and generate graphs and reports etc. However, AI is used also in background data processing automatically even thought that people are not even aware of the different AI layers.

AI history examples

Artificial intelligence took first steps in the automotive industry somewhere 2007 when the first DARPA urban challenge competitions were held in California. Of course, sort of intelligence has been part of automotive industry ever since when first sensors and CPUs were introduced in middle of 70’s. However, that was still far away from first neural networks were piloted in industrial applications (product quality checking, etc.) in early 90’s even though that artificial neural networks (ANN) were first introduced already 1943 [1]. Due to lack of computation power it took more 35 years before first commercial applications were coming into the markets. Interestingly, the reason for introducing neural based computing has remain over the years: need to process huge amount of data.

In addition to physical road infrastructure, the digital road infrastructure started to become hot topic when the smart traffic lights and especially communication between vehicles and infra was introduced 2005 – 2012. The aim is to collect images not only from road cameras but also infra cameras and share the data in standardised format between road users and authorities. In 2015, the research community was even talking about physical internet where parcels are divided to small pieces and routes are optimised like the TCP/IP packages in internet. Package delivery and route planning is one of the most traditional task for artificial intelligence due to huge number of options and parameters to be optimises.

This data economy is one of components and motivates to utilise more intelligent algorithms to convert data to valuable format. Therefore, Google and many other companies were introducing map based services for mobility domain. Consequently, the companies and authorities which owns the data has been woke up to consider that the data is oil of smart traffic solutions and therefore, it has value. Tesla is one of the pioneers in data processing from their vehicle fleet. They have an idea that 1) collect data from the vehicle, 2) process the data to more valuable format and 3) sell the processed data back to vehicle owner.

Smart mobile car

Navigation systems were introduced for wider audience somewhere middle of 90’s when the first handheld devices were coming to the mass markets. The first navigators did not use AI but soon when the maps become more versatile including various attributes to be updated continuously. There were manual updating centres where people and also map providers send updating requests for manual process. Soon started to consider if massive amount updates could be automated and also route planning can enclose not only shortest route but also take into account fuel consumption and congestions in the route. The only sensible way was to implement artificial intelligence for optimising non-linear optimisation problems.

The major expected trends caused by penetration AI based systems to the markets:

Artificial intelligence requires high computation power:

  • Automotive industry is partnering with chip providers to have computers on board (e.g. co-development of Mercedes-Nvidia in Europe)
  • Public transport operators are investing both to own and cloud based servers to process data more or less real-time. Otherwise the data pipelines get stuck
  • Big players (like Google, IBM) are really looking the opportunities to use quantum computing technique for dedicated purposes in background
  • Negative by-product is that server halls and CPU are already today causing about 2 % of world carbon dioxide emission and due to bigger server rooms this expected to raise to 5-6 %.

Capabilities to predict potential challenges from the data:

  • The big data amounts enables to utilise mobility data to predict congestions and provide detours in advance. Today, the prediction is done for couple of hours ahead but in future, travel times are more accurate even weeks before
  • More intelligent route planners for public transport where combination trams, busses and shuttle service are merged with automated robot taxies enabling real door-2-door transport.

Distributed AI system:

  • The AI is localised to respect privacy rules and provide local services. The distributed intelligence has hierarchic nature : single hand-held device <=> vehicle <=> edge computing unit <=> cloud based computation. Single device is personalised to user behaviour whereas cloud based system has generalised features
  • Distribution provides an opportunity for mobility stakeholders to provide local and personalised services (including safety warnings).

Privacy rules:

  • Its’ becoming mandatory that also AI based solutions follow the privacy rules regulated by the European Commission. This prevents some of the application to collect personal data including vehicle register plates. On the other hand, protection of personal data also limits personalising the services according to user needs. There is also danger that the U.S. companies will dominate the AI data markets and Europe is over-regulating AI approval.

Benefits of AI in automated driving

Automated driving functions

SAE introduced the definition for different automation level starting from 0 (no automation) and ending to 5 (full automation everywhere). Still automation level up to 3 (partial automation) is called more like a driver assistance systems instead of real automation levels. Its’ also slightly confusing what automation level 3 and 4 really means but nevertheless, 4 requires sophisticated systems where computer dominates and “driver” has only rare assistance role. When consider highly automated driving, the computers in the vehicle needs to process 1 Giga bytes of data in a second which is not possible with using deterministic algorithms and therefore, levels 4 and 5 are not possible without artificial intelligence recognising object and taking care of vehicle decision making.

Automated cars in a row

Sensor data

Modern vehicles equip with ADAS (emergency braking, lane keeping, etc.) bases on camera and radar technologies. The algorithms are trying identify objects in front with utilising information of speed, material properties and extracting specific features like lane markings or traffic signs. These are based whether assumption that metallic object detected by radar is front vehicle or other obstacle like fence which should be avoided. This works fairly well as long as driver is in charge of final decision making and automation functions has only assistance role.

Automated driving functions requires more intelligence and much more data not only for detecting object front but also recognise the object. Such an example is metallic can in front of the vehicle when hard braking is more dangerous for passengers than letting vehicle to drive through.

Trajectory and path planning

Selecting right trajectory in amazingly challenging problem. Typically, one overtaking of in motorway produces 5000 – 50000 different optional route alternatives. This depends on available free space, speed, accelerations, distance to vehicle in rear. The problem is even more challenging if taken into account non-visible lane markings for example due to snow or having gravel roads. Passenger cars are usually operating in the areas where there are special tracks, curbs or lane markings available. However, this is not the case for example to the machineries which are operating in forest or other open/closed areas. Selecting optimal route depends on many parameters like big rocks, distance between trees and softness of the ground. This is optimal case where artificial intelligence can have major benefits but only if the training data is properly selected. Usually, there are no deterministic one prevailing answer but rather alternative routes having pros. and cons.


C-ITS standards has been available already for 10 years. Europe expected already 2008 that connected driving is going to be booming next 5 years. That never happened even though that C-ITS has been deployed step by step but very slowly covering today backend public transport and some road maintenance services.

Artificial intelligence could the key enabling technology to gain implementation of C-ITS services. Now the biggest issue is that there are information provider and users but having very few stakeholders which converts the information to valuable format which users are willing to pay for. Taking an example – there are route planners to provide estimated time of arrival in different transport modalities and there are road weather services available. However, nobody estimates what is estimated impact of dense fog to different transport modes.

Mobility services

Behaviour of people has already significantly changed in the urban areas. Public transport is becoming more and more popular due to restrictions to use and park private cars in cities. However, public transport is not about big busses and metros like 15 years but today, they also include scooters, shared cars, bikes, Uber, etc. One of the challenges is that cost and route optimisation changes depending on location and there are people who don’t want to use for example scooters.

Artificial intelligent has two aspects to provide better mobility services:

1) more personalized solutions according to favoured transport modalities and willingness to pay

2) expand the city type alternative to inter-urban areas where e.g. automated shuttle busses could serve people when transport demand is high (e.g. rush hours in railway stations or some specific events)

Valokuva latauksessa olevasta linjaautosta


It's all about quality of training data. AI performance highly depends on training data quality. In mobility domain, the other important aspect is coverage of the training data. Real driving on roads is far from ideal world. There are rare things like tyres laying in the highway which happens maybe once per year but may cause serious unacceptable accident. The online artificial intelligent frameworks (e.g. ChatGBT [2] ) uses the information available online. Some of the systems are using cache to speed up the response time. Chatsonic [4] is an alternative for ChatGBT to process real online data. In mobility context, the ChatGBT biggest benefit is not probably the capability to produce reports but rather more interesting is the artificial intelligence for doing coding and software applications for mobility services.

Heikki Rajasalo
Heikki Rajasalo
Solution Sales Lead

Electric cars and urban cycling have become increasingly popular in recent years. Innovations are emerging at an increasing pace and technology is enabling solutions that are revolutionary.

Research expertise