The battle between Google and Apple is moving from smart phones to smart things, with both companies vying to provide the underlying architecture that networks your appliances, utilities, and entertainment equipment. Earlier in June, at its annual developer conference, Apple announced HomeKit, a new software framework for communications between home devices and Apple’s devices. Meanwhile, Nest, a maker of smart thermostats and smoke alarms that was bought by Google earlier this year for $3.2 billion, recently launched a similar endeavor with software that lets developers build apps for its products and those from several other companies.

Indeed, a quick look at the “Works with Nest” website reveals just how interconnected our future is about to become, with smart cars telling our smart thermostats when we’ll be home, smart dryers keeping our clothes “fresh and wrinkle-free” until we arrive, and household lights that flash red when the Nest detector senses smoke or carbon monoxide.

In fact, though, many of us are already living amongst an Internet of (some) things. We have desktops, laptops, cell phones, streaming devices like Apple TV and Roku boxes, and even smart televisions. It’s just that these systems have barely begun to work together properly, and therein lies the problem.

The visions of Google and Apple will require a lot more than new frameworks and developer conferences to be truly transformative. They will require heretofore-unseen levels of reliability, security, and usability. Otherwise we’re in for a frustrating and possibly dangerous networked future.

Wi-Fi is a key enabler of the networked home. But while Wi-Fi is now present in more than 61 percent of U.S. households, many homes have incomplete coverage, and when Wi-Fi doesn’t work, debugging is difficult. It will need to be dramatically more reliable than today to support the networked future.

Broadband Internet will need to be more reliable as well—as reliable as electric service is today. For many this may mean cable modems that can fall back to some kind of wireless 4G service, perhaps from a different provider. These modems will need to be dramatically easier to install and maintain than today’s.

We will also need improved debugging systems for when the Internet doesn’t work as it should. Today the primary recourse when your Internet is down is to reboot the cable modem, the laptop, or the smart TV—or even all three! And perhaps the problem wasn’t even in the house. To legitimately be considered smart, smart devices must assess what’s wrong with the connection, and then help fix it.

Connecting anything to a secure home Wi-Fi network is a challenge for many. And some devices need additional authentication information, such as an Apple or Google username and password. When passwords change, the smart objects need to get the new passwords, or they cease to work.

This approach of binding our smart devices to our personal accounts may be an easy engineering decision today, but it will make less sense as more devices show up in households with multiple family members. Families shouldn’t be forced to decide if the dishwasher is bound to Mom’s Gmail account or Dad’s. Instead, the household should have its own identity, with different family members having different levels of access depending on their needs.

Differential access will also be critical for the wide range of formal and informal arrangements that many households require. Think about babysitters, housecleaners, maintenance workers, and building superintendents. If these people need some way to interact with your smart devices, there should be some way to give them that access without sharing your username and password. And there should be some way to review their actions after the fact. And all of this delegation and auditing will need to be easy to configure and use without reading a manual or watching a video.

Beyond the issue of usability, the smart home will be an attractive target for hackers and malware. Even if the devices themselves repel attackers, other points of vulnerability include malware-infested desktops, laptops, and mobile phones. Smart things will be attacked, almost certainly in ways that we can’t anticipate today. Even simple data leaks might cause significant problems if they can be systematically harvested and exploited—for example, thieves might be able to determine when you’re not home. Voyeurs might hack your surveillance cameras.

With both Google and Apple aggressively moving into this space, another concern is the degree of compatibility between devices. Today, these firms are erecting barriers between their home entertainment offerings, with Apple TV and Chromecast, for example, offering separate content, pricing, and streaming models.

Some third-party vendors will surely try to stay out of this fight, offering apps that run on both iOS and Android, or are simply controlled via a Web interface. While that kind of strategy might work for a smart light bulb, it’ll be harder for the maker of a major appliance. If companies chose one ecosystem over another, it will be hard for consumers to switch from Apple-powered appliances to Google-powered ones.

Two things about the smart home of the future seem sure. First, given the array of resources being lined up on both sides of this fight, there is unlikely to be a dominant winner, meaning less flexibility for homeowners. Second, the coming wave of smart devices will rely on technology that is ill-equipped to guarantee reliability, and will also introduce completely new ways for things to go wrong. So the companies that make them will need to put far more focus on security, usability, and privacy to earn both customer acceptance and trust.