In 2002, when I was working at the Bioterrorism Preparedness and Response Program at the US Centers for Disease Control and Prevention, we noticed something unusual: In China, the price of garlic — which is viewed as a cure-all by many Asian cultures — had risen tenfold. This was the first sign of SARS, the predecessor to Covid, and we spotted it five months before China revealed that it was experiencing an outbreak of a new respiratory coronavirus.

In late 2019, satellite images from China showed cars were heading to hospitals more often than usual. In early 2020, the price of garlic rose again. These were among the early signs of Covid-19.

Often, we see the signs of a new outbreak before we see people getting sick. These early warnings can be invaluable if we learn to pay attention to and act on them. But they are not enough — we need to incorporate them as part of a larger system that synthesizes information from many different sources to enable us to spot new diseases faster and curb their spread sooner.

We need, in short, an immune system for the planet — a network of tools that search for signs of new infections, directly detect and analyze new pathogens right when they appear, and identify, develop and, ideally, deploy effective therapies (and perhaps vaccines).

The “immune system” would continue to rely on existing tools, such as monitoring demand and prices for medicinal therapies as well as satellite images of traffic patterns, and step up our efforts to monitor public sewers, as  pathogens often first appear in wastewater. But it would also incorporate additional strategies that enable us to intervene much earlier in the life of a new pathogen. These tools would look for novel pathogens in the air, water or soil, sequence their DNA or RNA, then use high-performance computers to analyze the molecules and search through an index of known therapies that might be able to neutralize the pathogen.

These early warning and detection tools could consist of biosensors that can detect pathogens embedded into organisms such as animals and plants living in tropical regions rich in biodiversity, where new infectious diseases often originate. Transmissions from these sensors would link to a supercomputing network that characterizes new pathogens. Not all detected microbes are harmful, so the system would have to develop computational models based on known pathogens to predict whether or not newly discovered microbes could cause problems for humans.

Once a harmful pathogen was discovered, the supercomputer could scan libraries of existing treatments that might fight it. Ideally, the system could one day automatically produce a customized mRNA vaccine that targets the pathogen, like some of the Covid-19 vaccines currently being deployed.

This future isn’t as far away as you might think. This past year, supercomputers played a key role in the fight against Covid-19, illuminating many details about Covid and how the virus behaves, including how long droplets can linger in the air, and pinpointed a gene ( F8A2) that may partly explain why some people may be more vulnerable to Covid than others. Research teams using supercomputers have even identified two existing drugs that might fight Covid, after scanning a library of about 1,600, as well as compounds in medicinal plants that may have antiviral properties.

Once future supercomputer models identify potential treatments to quell a future outbreak, they can help identify the best way to respond — either via conventional means, such as oral medicines or vaccines, or more experimental approaches. To distribute treatments quickly and equitably, it may one day be more efficient to add them to rain or drinking water, along with chemicals to time-limit the therapy. This, too, isn’t so far-fetched, at least within a 5- to 10-year time horizon: Several US states are trying so-called cloud seeding — in which they add small particles of silver iodide to clouds — to boost rainfall. Most Americans drink  water with added fluoride, and  cereal grain products are often enriched with folate. These automated solutions take humans out of the loop, and, in the case of Covid, could exponentially reduce the time it takes to mitigate a biothreat, thus saving lives, property and national economies.

Of course, such a dramatic scaling up of monitoring and therapeutics could raise concerns about privacy and personal choice, so we will need to take steps to ensure this planetary immune system doesn’t create a surveillance state. One possibility is to start with a pilot, perhaps monitoring future outbreaks of Covid in one community that buys into the idea. The project would be completely transparent, specifying standards for the ethical use of data, with perhaps a citizen group that oversees all decisions. If the pilot works and people feel comfortable, it could expand to other communities.

If this kind of global monitoring system works for disease outbreaks, it could be used to tackle other global health concerns, such as food shortages. Why not use artificial intelligence to develop indicators, warnings and plans to spot vulnerabilities in global food production, such as wheat, and help make the system more resilient and sustainable?

What I’m proposing is a massive undertaking, I know. But just imagine if something like this had been in place during Covid — how much suffering and loss could have been avoided. Covid isn’t over yet, and we know the next pandemic is coming — we just don’t know when or from where. If we start small, working in one area and expanding from there, we can show that this planetary immune system can work, and keep us safe.

This article is part of  Reset: The Science of Crisis & Recovery , an ongoing series exploring how the world is navigating the coronavirus pandemic, its consequences and the way forward. Reset is supported by a grant from the Alfred P. Sloan Foundation.