Google has long sought to diversify its revenues streams away from search and advertising, the business it was founded on and which continues to make up the bulk of its revenue nearly 20 years later. So could health be the industry that helps the company to achieve that aim?
Like Apple, Google has made a play for the lowest-hanging health fruit, the consumer market. Its first attempt to enter the market was with Google Health in 2008, a project that was meant to let individuals gather their personal health data and medical records inside a single Google store.
It was shuttered at the end of 2011, after failing to gain any significant takeup. Three years later, it announced Google Fit, its equivalent to Apple's HealthKit, which was launched around the same time. Like HealthKit, Google Fit is a platform for developers to build apps on -- apps that allow users to track their exercise or wellness data through an Android phone or tablet, or a Google Wear wearable. Google doesn't release user numbers for Fit, so it's impossible to say if it's proving more successful than Health.
Elsewhere, Google has a handful of other products designed for professionals working in the life sciences ad and healthcare industries: cloud services that allow genomics researchers to store, process, and share big data sets, for example, or a HIPAA-compliant version of its G suite for use in healthcare organisations.
While these are versions of its standard-issue cloud software given a different skin for the health industry, its most interesting work in healthcare involves products that have yet to see the light of day -- and could be revolutionary when they do.
Looking to the future
Google is making its biggest bets in healthcare through three separate units: DeepMind, Verily, and Calico. The two first two businesses are both building software and technology aimed at healthcare providers, the latter is ostensibly trying to defeat death using bioetch.
Of the three, DeepMind is perhaps doing the most familiar work. DeepMind, an AI startup founded in the English city of Cambridge in 2010, was bought by Google four years later. While DeepMind is not exclusively a healthcare company, its products with the clearest path to commercialisation are focused on the industry.
Its best-known healthcare product is Streams, an app designed to decrease the incidence of acute kidney injury before it occurs by alerting clinicians to the warning signs that indicate a patient is a candidate for such an injury. The app itself doesn't contain any AI at present -- think of it as more simple analytics software for healthcare -- it's likely that such elements will make their way into the products in future. The system is being trialled with the Royal Free hospital, and may be extended to other conditions where picking up the right signs early on can prevent a full-blown life-threatening condition, such as sepsis, taking hold.
Other partnerships with UK healthcare organisations show the direction of travel for DeepMind's products. For example, in pilots with the Moorfields Eye Hospital and University College London Hospital, DeepMind is testing whether its products can analyse medical scans more quickly than doctors can, cutting down the delays in preparing a patient for surgery or another procedure. If the pilots prove successful, DeepMind can sell the software as a means of cutting down doctors' busywork, so they can get on with seeing and treating patients - that is, the all-important stuff that technology can't do (yet). Few hospitals would be able to resist the lure of a system that can automate mundane tasks so clinicians can get on with work that requires the human touch.
It's a question of when, not if, such systems from Google and its competitors make it into secondary care. The only problems likely to hamper full-scale rollouts are cost (particularly in the UK) and patient privacy - an area that DeepMind has drawn criticism for in the past.
While DeepMind's activities are relatively well-publicised, thanks to its work being in the public sector and its publication of research in scientific journals, the same can't be said of Verily.
Verily's interests are more diffuse than DeepMind's (so far), and have a greater focus on hardware: Verily has teamed up with companies including Nikon to work on retinal imaging systems, Johnson & Johnson on surgical robots, and Dexcom on glucose monitors.
In general Google is happy to leave the hardware to other manufacturers so long as they are spreading its influence: most Android phones are made and sold by the likes of Samsung and LG, while Gear hardware is also largely in the hands of third parties.
Verily looks likely to pursue the same strategy: let another company make the hardware and focus instead on the various software underpinning it. It's this strategy that would seem to be the thinking behind the diverse interests Verily has: the likes of Nikon, Johnson & Johnson, and Dexcom will bring the hardware elements as well as the healthcare pedigree, while Google will be able to provide the software underneath. If it can make an OS that works across a range of medical devices, in future, it could turn that OS into the Android of healthcare devices -- a market thought to be already worth nearly $400bn, and one that's set to grow exponentially.
Not all of Verily's partnerships are such straightforward unions with healthcare companies but even its apparent curveballs fit broadly into the same pattern. Take its recent project to release 20 million mosquitos around Fresno, for example. By releasing only sterile males carrying a particular virus, the aim was to reduce the number of offspring from the female mosquitos in the wild and the rates of mosquito-borne Zika as a result. However, the breeders of the sterile mosquitos needed to sort them into male and female, and only release the former. Verily has developed a custom set-up using computer vision tech that enables that sorting. Viruses carried by mosquitos are responsible for several million deaths a year, according to the World Health Organisation, and hundreds of millions more infections. If Verily can crack the software that underpins a system that reduces such diseases, it could potentially take a slice of a very lucrative market: hundreds of billions of dollars are spent on malaria control in Africa alone.
As such, both DeepMind and Verily fit a mould that Google has created and exploited with other business units before. Calico, however, is a very different proposition.
Calico's stated mission is "tackling ageing". According to its own website, the company wants to "devise interventions that enable people to lead longer and healthier lives" by better "understanding the biology that controls lifespan". That is, it views ageing as a pathology which it can pull apart, understand, and defeat - not only beating age-related diseases, but death itself.
Since Calico was created in 2013, it has put out 15 press releases and its execs generally don't give interviews.
Apart from a few announcements about partnering with a handful of organisations on drug development work (sample titles: 'C4 Therapeutics and Calico Enter Strategic Partnership to Discover Novel Therapeutics Based on Targeted Protein Degradation' and 'AncestryDNA and Calico to Research the Genetics of Human Lifespan'), how Google's spinoff aims to achieve its ambitious goal is largely unknown. Calico says it's because its working on timescales of 10 years or more for products, so it has nothing to share in the meantime; others say keeping tight-lipped is just how Google does business.
As a result, it's hard to say just how Calico fits into the broader Google gameplan. That said, with the number of people aged 65 or older across the world is expected to rise from 524 million in 2010 to nearly 1.5 billion in 2050, it's clearly addressing a growing market. Thanks to advances in hygiene, vaccinations for infectious diseases, and safer childbirth, more people are surviving to grow old, and so are now living to develop conditions that chiefly affect the elderly.
Drug discovery is already being revolutionised by the advent of AI and machine learning; it's not unreasonable to think that if Calico could cherrypick from Google's array of homegrown tech - DeepMind's AI, Google's own ability to store and process eye-watering amounts of healthcare information - as its own top talent and partners, it could identify the biomarkers associated with age-related disease that it's questing for. And, once they're found, Calico and its partners could potentially begin work on arresting or reversing them. From there, could delaying death be next? The payoff for victory in either endeavour imaginable.
With such little information about Calico's work available, it's difficult to make an assessment on its chances of success. The one key advantage that Calico has in its favour is Google's backing, and the cash and gravitational pull that goes with it: it started life with a $500m funding injection from its Mountain View parent.
Even though Calico isn't planning to deliver any products for decades yet, Google can continue throwing money at the business for any number of years to come, as long as its search-and-ads juggernaut continues to pay out. Healthcare may be Google's future, but it's a future that's very much underpinned by its past.