Toyota has put $1 billion (£663 million) aside for a Silicon Valley artificial intelligence and robotics research unit to further its driverless car technologies, but is it enough to take Elon Musk's crown as the biggest tech spender in the car industry?

The Japanese carmaker's new facilities will open January next year, with 200 employees based across two locations - MIT University and Silicon Valley.

Driverless cars are ready to roam © iStock/Jason Doiy
Driverless cars are ready to roam © iStock/Jason Doiy

It's hoping to discover new materials and methods for manufacturing cars more efficiently, as well as creating safer, more accessible car technologies with the help of some of the brightest computing scientists in the world led by a former DARPA employee, Dr Gill Pratt.

The news will be certain to get Elon Musk’s attention. His electric car company Tesla already uses machine learning for its auto-pilot and lane changing features and is rapidly patenting any breakthroughs so his peers can't catch up. 

But Toyota’s AI newly announced commitment at around $280 million per year over the next five years still falls short on spending next to their luxury competitor.

Tesla spent an eye-watering $527.7 million (£349 million) between January and September on research and development for materials, development and testing of driverless features.

Its expenses began piling up when security flaws were spotted in Musk’s fleet of supercars, forcing the Tesla team to work on new software and auto-update cars on the road.

Artificial intelligence: Why is Toyota investing in machine learning for cars?

Foreseeing an ageing population, Toyota will focus on combining AI and big data to create cars that can assist mobility both in and outdoors for those less able to drive.

Being first in the race to create safe, effective autonomous driving systems is paramount for carmakers, which have seen profit losses as younger generations move to service models - favouring Uber and the sharing economy over car ownership.

Toyota’s general research and development budget is still impressive - and much larger than its European peers. It worked out at around £2.4 billion in the first six months of last year.

Audi: ‘Driverless cars will allow the car industry to become an innovator again – but we cannot do it alone’

Last week Audi’s head of digitalisation, Sven Schuwirth took to tech conference Web Summit’s stage to implore startups and techies to help make cars relevant again.

The German carmaker is researching how to use machine learning and robotics with the goal of giving customers an extra 60 minutes usually spent travelling.  Audi believes the extra hour translates to “a hell of a lot of money”, Schuwirth says, for both car owners and the wider economy.

While car owners lean back in their back seats, it’s likely they will be using devices that will serve them advertising content, providing new revenue streams for other industries. Last year Audi spent about £1.4 million on general research and development between January and June.

But Volvo, which is already trialling driverless cars on the roads of Gothenburg, spent around £1 billion between July and September.

It’s not just carmakers that are siphoning R&D budgets for AI. Applying deep learning methods to big data can produce smarter products and find out more about customers. This is particularly valuable for advertising, retail and general manufacturing companies.

Volvo's driverless car system is packed full of sensors and LiDAR tech to sense its surroundings ©Volvo

Perhaps the most well known example is Google. The Silicon Valley heavyweight’s most profitable business, Search, is based on artificial intelligence that serves recommendations and targets users for advertising. It has been building proprietary algorithms since its inception, but has been beefing up its AI business portfolio lately.  

Last year it bought AI startup DeepMind for £400 million. DeepMind, which was the brainchild of a former chess prodigy and based in Cambridge University, has moved into Google’s London offices and is working on creating the most sophisticated methods for machine learning.

Next, Google splashed its cash on Germany’s top AI research institute and recently backed Beijing-based Mobvoi.

It took a minority stake in the company’s £50 million fundraising round. The investment is allegedly part of the tech company’s mission to expand in China. Mobvoi develops voice controlled software which could be useful for Google Now or its Android OS in the Chinese market. Mobvoi was responsible for the Android Wear smartwatch OS.

Google also has its own self-driving car project, which already has 25 prototype Lexus models driving on the roads of San Francisco and Austin. It’s keeping the cost of the project under wraps, but the company spent almost £6 million on research and development between January and September this year, and it’s likely the AI element of the car software would have taken a chunk.

Artificial intelligence in business: connected boilers, customer service and crime

Tinder, Twitter, Facebook and Instagram are other examples of businesses built on deep learning technology, but there are several pure play firms making the leap too.

Some companies are using IBM Watson’s supercomputer to bring AI into to the workplace.

Imperial College London is helping to predict crime with the Watson team to develop a cognitive computer app to solve mysteries brought to the university’s crime department’s attention.

African Standard Bank uses IBM Watson to speed handling of customer queries, allowing it to identify customers quickly so they can respond in faster time.

Pharma company GlaxoSmithKline has created a disease mapping tool with Watson too. It uses clustering techniques similar to that of Instagram and Facebook to determine relationships - but for genes and biological processes.

UK institution British Gas has mastered machine learning in the enterprise. The utility’s Connected Homes unit have created a predictive maintenance algorithm that will help alert customers when their boiler is about to pack in.

Similarly, the majority of smart homes and internet-of-things related products will be underpinned by artificial intelligence. Thankfully, they’ll be less like HAL and more like Hive. For now.