Why It's Too Early To Write Off The Internet of Things The Internet of Things (IoT) used to be the talk of the town. While its promises have not panned out, IoT may still have a bright future.

Just a few years ago, the Internet of Things (IoT) was the talk of the town. The promise of an interconnected web of objects equipped with sensors of all stripes that would communicate with each other in a way hitherto only envisioned in science fiction was seen as imminent.

Startups sprung up left, right, and center, and segments flooded the airwaves about new smart cities, like Saudi Arabia's sci-fi-inspired project, Neom. That reality has still not panned out, and it still seems distant. Some have written off IoT and almost thrown it in the dustbin of history along with other defunct technologies that didn't deliver on their promises.

One of the challenges IoT faces is that it depends on a very high level of high tech to be deployed in many places. A smart system that recognizes a light bulb needs fixing will require some sensor that communicates with each light bulb.

You can't just plug devices into a "'smart solution" and suddenly expect a technological nirvana to unfurl. In an incredibly technologically heterogeneous world covered with both high and low tech, you need integrative technological solutions, which are still rather limited.

Related: Is Your Business Ready for the Internet of Things?

The world can, however, integrate existing solutions much better. Positioning systems, cybersecurity systems, and physical security systems already exist. Most necessities are on the grid, and everyone is online. We can put together the tremendous technological tools already at our disposal by utilizing integrated technologies and get a long way toward IoT's promised land.

Some creative ways exist to bridge the need to equip the world with sensors. Elon Musk realized this when he argued that self-driving cars do not need to be equipped with radar. Instead, they can rely on sight just like humans and still possess superhuman driving skills. We, humans, use sight to identify objects, events and threats at a distance from us. There is no reason why machines could not utilize vision as adeptly.

Computer vision extends far beyond the novel wonders of self-driving cars, possibly even into every little thing about our lives. Data-annotation service provider Keymakr, for example, recently joined forces with SeeChange to leverage AI to reduce the number of times shoppers and employees slip, trip, or fall in brick-and-mortar stores. The AI identifies and notifies employees of liquid spills in fall-risk areas.

Computer vision in this scenario prevents stores from having to equip the floor with additional sensors to detect if it's slippery, instead using cameras already in place. Imagine the boundless other applications for such technology, ranging from predictive maintenance to reshaped hospitality with automated services or a new level of proactive and personalized remote healthcare. The potential applications are bound only by our imagination.

We will have to address the issue of security, considering by now, we have the experience to know that almost every device is hackable. Connecting all the world's devices poses brand-new security risks. We all read about exposed personal data hourly and experience too many technological failures daily. Are we ready to trust a vast network of integrated electronic devices to run the world smoothly and safely?

After all, IoT devices run on software susceptible to many vulnerabilities that can be exploited. As more and more devices become connected to the internet, we will face an increased risk of hackers accessing data gold mines from massive networks that were previously much more challenging to target. They'll do so by attacking less secure IoT devices connected to that network.

Focussing on individual vulnerabilities, however, won't yield the most effective security outcomes. Instead, it results in a much more costly, computerized version of whack-a-mole where the security professionals run after vulnerabilities to patch them up one by one.

By taking a holistic approach to the security of IoT devices, cybersecurity company Sternum IoT builds itself into the system's firmware to ensure the code can't be tweaked. Simply put, even if a malicious attacker could hack into the device, they would be barred from actually performing any of the functions that inflict harm.

We need more proactive takes on IoT security to ensure companies can come out ahead instead of playing catch-up with hackers and constant costly vulnerability patching, as security is usually performed today.

IoTs' promise to truly connect us and technology in a new way is similar to what's happening with self-driving cars. We heard all about it constantly for a period, and one could be forgiven for thinking we'd all be driven around by machines by 2023.

While the technology is still not ubiquitous, it is advancing quite nicely. Think how much of the driving experience is already automated compared to just a few years ago. Cruise control, automated lane adjustments, and collision aversion technologies are only a few of the dozens of automated features.

Related: How Cloud Agnostic Hardware Could be The Future of IoT

With access to low-cost, low-power sensors, new levels of connectivity, cloud computing platforms, machine learning and analytics, IoT is already combining state-of-the-art technology into something new and exciting. It is certain that IoT will grow and that technologists will do well by staying ahead of the curve. But it remains to be seen how fast and for how long that growth will continue. It might just be that IoT is still like the sleeping giant which will move the world when it wakes up.