Fujitsu Changes Horses Midstream
With the tide of information flowing from Fujitsu in recent months, it is rather opportune that EITN was able to sit down with Craig Baty and Abe Chiharu at the sidelines of its Asia Conference, World Tour 2017 in KL recently, to suss out what’s the raison d’être for this 82-year old, Fortune 500 company and how it plans to “Shape Tomorrow With You” in the competitive years ahead.
Craig Baty is Fujitsu Ltd’s VP of Global Strategy & Digital Services while Abe Chiharu is CEO of Fujitsu Malaysia. The tide of information refers to Fujitsu Global’s announcement in January 2018 that it was selling down its mobile phone business to a 30% stake to Polaris Investment Group; as well as its announcement in November 2017, that Lenovo was buying a 51% stake in Fujitsu’s PC business. In that same month, Fujitsu Ten was renamed “Denso Ten” when Denso Corporation became a majority shareholder. The other major shareholder is Toyota Motor Corporation. Denso Ten develops & manufactures car audio, video, navigation and control systems.
Human Centric Innovation: Digital Co-creation
The theme of “Human Centric Innovation: Digital Co-creation” is perfectly elucidated as Fujitsu believes that technology is to be used to help make people’s lives better, therefore, putting people in the centre of everything.
The company does not fear the doomsday predilection of some academics and certain high-flyer tech entrepreneurs that AI-based intelligent systems and robots pose not only threats to job security but also an existential dilemma of sorts to humans! After all, A.I lacks “the rich tacit knowledge and intangible qualities, such as intuition, emotions and creativity” that humans have. Realistically, it will be for a long long long time that robots will lack “common sense”.
These questions and more were put forth and the following are the “sound bites” articulated by Baty and Abe. First off, instead of A.I being Artificial Intelligence per se, Baty re-termed it “Intellectual Automation”. After all, Fujitsu’s brand promise of “Shaping Tomorrow With You” is not meant to be to you, for you, by you, from you or without you… so the concept of human-centric intelligent computing is that Fujitsu works alongside “with you” to shape a better tomorrow and advance society, improve people’s work and lifestyle.
Typically, apart from its products, as a Japanese company, Fujitsu does not subscribe to a ‘global package’ for services, where every company must have the same standardised solution. It sees every company as different, especially in Asia. Every country has different rules and regulations. For a local client, if it comes up with an idea, Fujitsu will build something to make it work for that particular environment. Attention to detail is in the Japanese DNA.
To prove this point, Baty and Abe highlighted the following case studies or proof-of-concept (POC) projects undertaken not long ago.
“One of the problems in Russia is that in schools, children are not always buying healthy things at canteen time. So Fujitsu provides a system in certain schools where smart chips are embedded in food trays and after the children place their food choices on them, they go through a scanner. The online system records the children’s choices and so parents can be alerted,” shares Baty.
“The whole idea of Big Data, analytics and connectivity all culminates in being able to create solutions based on what customers truly need and not what we think they need, which means, the whole process may take some time. You see the example of Daihatsu Perodua in Malaysia, where Fujitsu built a system to optimise productivity in engine manufacturing from the ground up.
The reality today is that organisations will have new systems which need to work with old systems. After all, 75% of the systems are still on mainframes! Trillions of money had been invested in them, so it is impractical to get rid of those systems when they are still doing a good job. However, one might need to add something new, such as a new application programming interface (API).
Fujitsu does not subscribe to a proprietary system only – whether the customer chooses to upload onto a Fujitsu Cloud or Amazon Cloud or Azure; Fujitsu will customise the solution and integrate as necessary. For example, in a hospital mainframe system, the hospital need not exclusively use Fujitsu sensors,” continues Baty.
Future in Services and Software
Fujitsu is still a company that makes everything in the I.T space, but its future is in Services and Software. It is not giving up the “other stuff” (namely the PC and mobile phone sectors) but is looking at different ways to do it. Despite, taking a much lower stake, Fujitsu will still maintain control over the supply chain whilst moving its resources and funds to areas that are not necessarily high-growth, but is strategic.
It may not make a lot of money out of cloud nor IoT nor A.I. yet, but is heavily investing in these areas to build up its future as a leader. It considers itself one of the top companies worldwide with A.I plans since embarking on it from the 80s. It considers its A.I tech today as a 5th generation.
In Malaysia, there are some POCs being undertaken with city councils that although, in Japan, Fujitsu has ready tech solutions, customisation to Malaysian environment is required. For example, there is a private enterprise keen on fatigue management, i.e. sleepiness tracking. Based on climate in Japan, 31 degree Celsius will be flagged as a danger but it is normal in Malaysia. So factors like the climate, weather, body temperature, employees’ history, normal sleeping patterns all need to be studied and solutions adapted. On its drawing board is the plan to work with academia and recruit university students to provide reliable localised data to help adapt Japanese solutions to the Malaysian environment.
In neighbouring Singapore, there’s a company called Connected Health which incorporates Fujitsu’s IoT Solution known as Resident Monitoring Solution (RMS). This is a remote system which links a patient’s medical records to his health metrics and wearable devices. For example, in diabetes management, it allows healthcare providers to monitor and record glucose readings.
One of the challenges around the world, especially in Japan, is aging population and the exorbitant cost of healthcare. The best thing to do with hospitals is to keep people out of them. The best thing to do with doctors is to keep people away from them! It costs money not only for the patient but for any government with a national healthcare system.
Furthermore, for the disabled and the elderly, it can be challenging and time-consuming. Connected Life was set up to help families better connect with and care for relatives while respecting their privacy. People don’t want cameras watching them in their toilets or placed on their shelves. However, with motion detectors and sensors, it can sense when someone falls over. If an elderly who has dementia takes off his track-able wearable, the motion detectors in the room will still notice where he is and can monitor his body temperature.
Like a factory monitoring system, it can tell whether he has fallen over. Sound is a very good indicator. The system records the normal sounds of a person around the house as well as the normal movement patterns, and over a period of time, can tell if a sound or motion for that particular person is abnormal. The A.I can learn and when it discovers what is abnormal, it can activate and send a virtual doctor, or a family member or an ambulance over. So, Fujitsu provides the algorithm, the sensors and the A.I. for Connected Health.
The company is also working with Yamaha on an advanced sound system, which is at a POC stage. One can record the sound of the beach or of children playing or of a concert performance and later replay these and imagine being there, at a later time.
AKISAI and the Supercomputer
On a global basis, the United Nations has come up with 17 sustainable development goals (SDGs) and Fujitsu is strongly supporting these initiatives. Key areas for Fujitsu are agriculture, healthcare, transportation, finance, drug testing development, communication and education.
Global production was one of the first things that Fujitsu started on as after the 2nd World War, Japan was devastated and there was mass nutritional deficiency (not enough protein, Vitamin C, carbohydrate, etc). In tandem with the government, Fujitsu developed an agriculture cloud service, called AKISAI. Sensors were activated to monitor soil and weather conditions, cameras & sensors were inserted on tools and tractors, even on farmers’ uniforms and record when they did their maintenance.
All these information were loaded onto a database and became a knowledge base on how farmers ploughed and when they ploughed. This same knowledge was sold back to the farmers, who can make use of AKISAI’s smart algorithm that can connect them to weather forecasts so they can assess if the weather was good enough in 3 weeks’ time to plant the seeds that would yield 2 tonnes in a few months’ time. Today, AKISAI’s algorithm can also work out that if the farmer waits a week, he will only yield 1.5 tonnes. This connects to the backend system, so if the farmer is going to plant 10,000 seeds, the supplier will be alerted to deliver 10,000 seeds. And it tells the government to anticipate 2 tonnes of rice in the next few months.
Increasing urbanisation means eventually, fewer of the farmer’s next generation are keen to farm, forcing the farmer to hire outsiders. The farmer must now deal with HR rules, insurance, occupational safety, scheduling, etc. AKISAI has a built-in farming system which provides the HR system, the scheduling system, that all comes in one package.
There are 10,000 used cases that been very successful in Japan, it has gone ’live’ in Vietnam and at POC stage in Australia with a very big organisation. The trouble in Australia is that the first sensors can only send information a few hundred metres as there are farms in Australia that are as big as Texas or Singapore! So Fujitsu is still working on sensors with a bigger range.
Sometimes the tech that works in a country may not be workable in another. But that is just part of the Co-creation ambit where Fujitsu aims to build the solution together.
More than a decade ago, Fujitsu’s supercomputer, touted the “K Computer” was developed with RIKEN (Japan’s leading research institute) and when it went ‘live’ in 2012, it was the world’s fastest computer. But quickly it became like an “arms race” where every other high-tech organisation began touting its supercomputer to be a faster computer… except that they are all for very specific measurements. The interest is because everyone recognises that “data is the new oil”!
The reason Fujitsu is so invested in the K computer is because it will power A.I and enable faster critical decision-making. Fujitsu Intelligent Society Solution Smart Mobility (SPATIOWL) has been used in Japan for earthquake predictions and modelling. It can deal with large amounts of data coming from multiple sources such as public transportations, vehicles and pedestrians’ smartphones in urban areas.
A year before the 2011 Fukushima earthquake, SPATIOWL was being tested as a POC. When the tsunami hit, in Tokyo, all the train stopped, but the systems were still running and could track where all the taxis went. K Computer was able to simulate and tell the first responders the best routes to take using real live data. Because Fujitsu had placed thousands of sensors floating off the coast of Japan, with the rise and flow of the waves, they could sense the sand particle movements and where schools of fish went.
During the POC stage, 4000-5000 Tokyo taxis were outfitted with sensors on windscreen wipers – sensors sensed the wipers moving rapidly when it rained, predicting that more taxis would be hailed. For traffic, there were sensors on the ABS brake systems too. When sensors sensed repetitive “stop start stop start”, such data together with the GPS coordinates was sent to the traffic control centre.
Cameras and sensors put at bridges and toll booths enabled massive amount of data to come in – which helped taxi companies plan routes to bypass traffic jams and cut carbon emissions. Predictive modelling also meant traffic controllers could decide when to open a second lane on the bridge to ease traffic or even if they notice repetitive accidents at a certain location, they could check if the road there needs fixing.
Going forward, this sort of data can help predict what will happen when a tsunami next hits. The supercomputer can do predictive modelling within 45 minutes, if a coast will be hit and how high the waves will be. If we can pinpoint within 2 kms of where a tsunami will hit, we can save all that evacuation costs & time of other unaffected areas. SPATIOWL, tsunami modelling and climate modelling are what Fujitsu is doing with the K computer.
More customer stories and case studies can be gleaned from the Fujitsu Global website.