FORTUNE — Hurricane Sandy revealed the vulnerability of our power and transportation systems. New York City, one of the world’s major economic centers, was brought to its knees by a storm that we knew was coming. The NYSE and Nasdaq exchanges — the lifeblood of world markets — were closed, despite all their investment since 9/11 in backup systems. More than 8.1 million people lost power across ten states. More than one hundred people died in the US.
How can we upgrade our cities so that they function during such adverse conditions? Where should we invest in infrastructure to get the best return? Answers may lie in the design of another crucial system: the Internet.
The internet was built to be resilient — it was engineered to keep functioning even if part of the network went down. The designers of ARPANET, the precursor of the modern internet, chose a distributed system with no central node for their network. They found ways to route information in multiple paths from origin to destination to avoid a single-point of-failure fault. This is one reason why the Web is so robust today, even when it is attacked by hackers or its physical elements by natural disasters, information can be re-routed around the fault to reach its recipients.
Unlike the internet, our power grid is architected such that if any fault takes place along a line, all customers on that line lose power. Steven Johnson calls these “Legrand Star” networks in his new book, Future Perfect, after the central-node railway system in France. It turns out that our power grid is even more fault-prone than a central-node network. In 2003, 45 million Americans lost power due to a sudden blackout. In that disaster, the entire Eastern half of the US lost power when just one utility plant in Ohio went offline. What would happen if hackers tried to take down the system at multiple points?
And yet, we learned little from that catastrophe and did seemingly nothing about it. Our grid is the very definition of the fragile system that Nassim Taleb warns against in his new book, Anti-fragile. Taleb urges us to redesign systems to learn from failure points and get stronger.
Our grid today is just as fragile as it was in the blackout of 2003 — maybe even more so. The grid that serves lower Manhattan was the first utility network in the country, built by Thomas Edison in the 1890s. It’s time to upgrade the infrastructure in NYC and across the US. Here are specific steps we can take to increase the reliability and resiliency of our electric grid:
1. Distributed Electric Storage
Our power system today cannot store the electricity it produces. Once electricity is generated at a plant, it must be used immediately or has to be grounded into the Earth. When it is not used and not grounded, high-capacity power lines heat up, sag and start faulting due to interaction with trees and other ground-based objects This is in part what caused the 2003 blackout. Since there is no storage of any scale in the utility grid, there is no buffer space when any kind of fault occurs and the system quickly shuts down in a domino effect.
Imagine a computer network without hard drives — a network where you had to use all data produced in real-time and could not store or retrieve any of it. That would be ludicrous, but it is the very system we have in our power grid.
Storing the electricity we produce in the grid equivalent of hard drives was historically quite expensive and power authorities wanted to avoid any project that would raise customer rates. The good news is that electric storage is now getting cheaper and the first commercial storage units are being installed this year across the US.
These batteries can pay for themselves in a few years by enabling buildings and homes to use cheaper, off-peak electricity. In case of a natural disaster, they also provide backup and line-conditioning to smooth out any spikes which can fry appliances and electronics. Battery storage gives us these benefits without the use of gasoline or diesel fuel.
Presidio, Texas is a town that decided to install battery backup to its grid since it is tied to the rest of the Texas grid by only one line. A joint venture of AEP AEP and Berkshire Hathaway’s Mid-American Energy brk.a installed the battery units.
One of the reasons the internet is so robust is that it has both large data centers in the cloud as well as distributed computing in every users’ laptop or tablet. This combination allows any user to be productive no matter where they are and whether or not every network node is working or not. Amazon (AMZN), Goolge (GOOG) and Apple (AAPL) benefit from this robustness and can scale their businesses easily with both cloud-based services and local apps that users have on their own devices.
While the power network has centralized generation in the form of large-scale power plants, it has limited ability to generate power at the edges where the users are. This creates more points of failure instead of a resilient system that can adapt to a sudden blow to one part of the network.
Distributed generation includes a range of technologies from natural gas in a combined heat and power setup to renewables such as geothermal or solar. Different power generation technologies can be combined right onsite. Having onsite power means that even if the grid goes down, the facility can still maintain emergency systems. This would have been of immense help for hospitals and apartment buildings that lost power during Sandy.
A number of companies supply the technology for onsite generation including GE GE, Capstone Turbine CPST and Bloom Energy. Bloom natural gas boxes are used by Google GOOG and other data center managers to provide immediate fail-over backup. A number of companies are working with building owners to assess their energy usage and options including Johnson Controls JCI, Jones Lang Lasalle JLL and CBRE CBE.
Dozens of countries already have distributed generation systems integrated into the main grid. The European Union, in conjunction with the Fifth Framework Program has invested about 70 million Euros in distributed generation systems, electricity transmission, energy storage, and high temperature superconductors for power supply in regional local and island grids. Eight leading European energy utility companies have joined forces for a massive deployment of DGS — we shouldn’t stay behind.
Our transportation system is also riddled with single points of failure that caused billions of dollars in productivity loss just in the last few weeks. Even though the NYSE and NASDAQ work with electronic trading they had to be closed for business since the traders were not able to make it to their offices—this is not a 21st century transportation system.
Here are three actions we can take to prevent the transport meltdown we experienced:
Add generators to filling stations
We should invest now in generators at filling stations to provide back-up power for our transportation network, which is dependent on our unreliable power grid. Filling stations are useless without electricity to power the pumps. Of all places, filling stations have enough diesel fuel to keep a generator going to power the pumps.
If gas stations want to hold off on purchasing a generator they can contract now with a generator rental company to supply one in case of disaster. Hertz Equipment Rental HTZ and United Rental URI are two such providers. (Full disclosure: I advise Hertz on EVs and learned about their equipment rental division – many building owners still do not know that these suppliers can outfit them with backup power.)
Convert streetlamps and all traffic lights to solar-powered LED units
LEDs and solar are both so cheap now that a solar-powered LED street lamp can pay for itself in less than three years with no subsidies. These units have batteries on board so that they can power themselves for days through any storm. LEDs last five times as long as current streetlamp technology which translates to big savings in maintenance. Los Angeles has already installed solar-powered LED streetlights and Mexico City has decided to do the same
Traffic signals should also be converted to solar-power with onsite battery backup. Dark traffic signals not only create dangerous driving conditions, they also require enormous police manpower to maintain order at intersections. Companies to look at in this space include CreeCREE, Philips PHG and OSRAM Sylvania unit of Siemens SI.
Invest in redundant pumps and power for the NYC subway system
We see what happens when the subway goes down — loss of productivity across the NYC economy as its daily 5 million commuters are stranded. The MTA is dependent on our outdated grid as electricity is needed to operate the third rail.
Even after the MTA had pumped out all of the water in the tunnels using diesel trains, following Hurricane Sandy, subways could not operate downtown because ConEd had still not restored power in those areas. To call for distributed generation and storage for the subwaymay be a tall order for a system that still has delays on a regular basis, but it is a worthwhile investment to build a more resilient subway network.
The MTA only has three pump railcars and should augment this with additional pumping and emergency power equipment. The MTA did learn from the 2010 blizzard the importance of moving cars to safe areas so at least those procedures have improved.
After Sandy, the board of Long Island Power Authority fired its CEO with good reason — he did not prepare for the storm and his negligence left thousands of customers offline well after other areas were back online. The story at ConEd will be different.
Let us not wait until we are hit with the next disaster that takes out our power and transport systems and costs lives and untold financial impact. Our negative real interest rates make long-term infrastructure investments with real returns an attractive opportunity in the current economic environment. Now is the time to invest in our infrastructure and to create needed jobs and value in the process. We cannot afford to wait any longer.
Jack Hidary is a startup investor and co-founder and former CEO of Dice.com . Send Jack comments @jackhidary.