This article originally appeared on Forbes.com.

Robots eliminate a lot of human error, but as with all technologies, they’re not perfect. Just ask General Motors, which uses thousands of robots to make cars. There was a time when the automaker learned of a downed robot only after the problem had halted an assembly line and created costly production delays.

GM recently found a solution in a tool called the Zero Down Time Application, which was developed using FANUC’s industrial robotic systems, Rockwell Automation solutions and Cisco’s IoT and cloud software and infrastructure. The software platform collects production data from GM’s robots and then analyzes it to anticipate repairs, order parts and prevent problems from turning into major disruptions. GM has deployed the system across 27 factories in five countries.

Such innovations are prompting more executives to ask: Where do we go next with IoT technology, and how fast can we get there?

We believe that over the next four years, the market for solutions from IoT and analytics vendors will swell to more than $450 billion. Without a doubt, many new devices will be wearables or app-based tools that are straight out of The Jetsons. But even more of these will be the kind of groundbreaking tools that help industrial companies detect product flaws using algorithms or let cities manage transportation systems with sensors.

Tech giants have been aggressively positioning themselves—many through acquisitions—for the inevitable land grab in enterprise and industrial IoT. Earlier this year, Cisco acquired Jasper Technologies, whose service platform helps companies manage their IoT services, for $1.4 billion. Last year, IBM bought The Weather Company’s weather data technology, allowing Big Blue to offer forecasting data to IoT developers, to integrate into their products.

Despite the excitement around the IoT, many tech companies are still in early days as they look for the best entry point into this complex space. As we stated in a recent article, we see the internet of things not as one market, but as five critical battlegrounds, each with its own opportunities, challenges and key players. Before jumping headlong into battle, companies need to know the dynamics of the battleground they are fighting in.

Consumer. Household names like Apple, Google and Samsung are out in front in this fiercely competitive market, which includes the wearable devices and smart appliances people tend to associate with IoT. Software and hardware companies whose platforms can gain widespread traction with consumers and solve current issues of interconnectivity will come out ahead.

Enterprise and industrial. Not surprisingly, this arena belongs to incumbent players. As industrial and workplace equipment and devices begin to connect—think smart factories and automated office buildings—incumbents will augment their own operational technology offerings to address IoT opportunities and form partnerships with analytics vendors.

Network and gateway. Whether it’s monitoring patients at a hospital or improving the customer experience at a retail store, IoT applications will increasingly depend on real-time services and edge analytics, creating opportunities for network equipment companies like Cisco and Ericsson. Meanwhile, telcos—AT&T and Telefonica, for example—can cash in on the sheer volume of new devices by providing authentication and location services and by expanding their software platform offerings.

Analytics. With the IoT projected to create 5 trillion gigabytes of data every year, you can expect analytics to be a critical battleground as companies mine that information for value. Traditional analytics vendors, such as IBM, SAP and Microsoft, as well as cloud service providers, such as Amazon Web Services, will be able to build on their existing customer bases to grow, while new start-ups can be expected to step in and provide more custom services.

Autonomous. Self-driving cars, drones and robotics differ from other IoT applications in some fundamental ways. For one thing, most of their sensor data is collected and processed locally (as opposed to being stored in the cloud and processed remotely). Encouragingly, this market is also wide open to newcomers, not just to existing industry leaders. Companies that master real-time technologies like machine learning and computer vision, and establish a reputation for safety will have an edge up on the competition.

Transformation of industries with IoT and analytics technologies represents one of the biggest business opportunities of our time, but the old rules of success still apply. Tomorrow’s top vendors won’t necessarily be the whiz-bang start-ups making the rounds with venture capitalists. The true winners will be those that address specific customer needs, invest in the right capabilities and talent, provide more turnkey offerings with partners and differentiate themselves in what will likely be a crowded market. And they’re starting to stake their claims now.

Ann Bosche is a partner in Bain & Company's San Francisco office and Michael Schallehn is a partner with Bain in Silicon Valley.