What Are the Tech Trends to Watch? Predictions From Aptiv Experts

Aptiv’s long history in the technology industry has taught us that the future is always changing. Technology trends come and go like fashion. But over time, the arc of innovation makes the world a better place for everyone.

As science fiction pioneer  William Gibson said, “The future is already here — it’s just not very evenly distributed.”

Through our experience in areas as varied as automotive, industrial robotics, aerospace and beyond, Aptiv has a unique perspective that helps us identify the trends that will matter over the long term. Below, five Aptiv technologists offer their best estimates of 2026 industry trends that will affect our customers and look beyond them for what comes next.

AI Will Become Even More Ubiquitous

The slam-dunk prediction is for an upswing in AI adoption — but what will that look like?

Technical advances are ensuring that machine learning will trickle into everything at the edge, says Sam Palmisano, Aptiv vice president, global product development, adjacent markets. One key development is sufficient embedded compute capabilities to support growing workloads. The hardware used to be expensive but has become more affordable, he says. It also used to lack the core processing “accelerators” needed to execute machine learning workloads. “The machine learning approaches in the past two to three years have changed massively with the introduction of transformer-based architectures,” Palmisano says. “You can [now] learn a lot quicker off less data, and you can get a lot more context off of that data.”

As a result, Palmisano says, machine learning can now be applied much more broadly.

For instance, early automotive machine learning was largely used for image recognition. Camera images would basically match pixel patterns to conclude “that’s a balloon” or “that’s a stop sign,” Palmisano explains. But now, in 2026, image recognition technology can use machine learning to correlate attributes and conclude that an object is a road sign. That capability is immensely powerful, and we will see the fruits of research in that area in the short term, he predicts.

ADAS Functionality Will Accelerate

In 2026, expect to see ADAS functionality improvements that resonate with OEMs and the users they serve. Concurrently, software development practices will continue to be adjusted to respond to those changing requirements.

Significant ADAS functionality improvements have been made in the past few years, enabled in part by radar, cameras and lidar, points out Brian Witten, Aptiv vice president of emerging technologies. Now is the time to convey their value. For instance, radar’s capabilities are well proven, the technology is strongly certifiable from a functional-safety perspective, and it delivers the necessary performance for full self-driving capabilities. In 2026, the technology’s merits will be borne out. “We already demonstrated prototypes earlier this year, and we look forward to maturing the technology with our customers throughout the year,” Witten adds.

This suggests a changing mindset for designers and engineers. Niheer Patel, Aptiv’s head of product management for middleware and DevOps, sees a shift in the software development process as OEMs contemplate workloads differently — one that is accelerating the shift from hardware-first to software-first. Historically, hardware selections constrained what the software could do. Now, Patel says, “engineers can design and develop applications before they select the hardware, including sensors, the [system-on-a-chip] — it could be a lot of things.” That will not affect existing functionality or vehicle drivability, but the change in developers’ approach can yield cost savings and quicker time to market for OEMs.

Physical AI Will Become Real

Artificial intelligence will expand what hardware can accomplish, especially in robotics.

“I don’t mean just more AI in the process, although certainly that’s going to happen,” says Doug Welk, Aptiv’s engineering director in cockpit software. “Construction cranes and things that are relatively basic today will get more AI functionality.” The AI might help reduce the workload for the equipment’s operators or improve safety for the people around it, he suggests. And it could help systems of equipment efficiently accomplish tasks with reduced oversight.

For instance, to prepare a plot of land for a new building, a single operator could instruct excavators, dump trucks and bulldozers. The excavators would know exactly where and how deep the trenches should be; dump trucks could deposit the fill exactly where the bulldozers need it in order to build a berm, and so on. Even the fuel truck could show up exactly when needed. AI could choregraph that dance perfectly, intervening only when something unexpected happens.

There are many use cases to explore. “As systems get more intelligent, they can do a lot of cool things that allow folks to become more efficient,” says Lexi Schroeder, senior manager for product portfolio strategy at Wind River. For instance, she asks, have you ever waited interminably at an airport’s baggage claim? While automation has improved the experience (such as sending travelers an alert that their luggage has been taken off the plane), most airports rely on humans to move and sort bags. That manual effort might instead be accomplished by autonomous vehicles or autonomous robots.

One effect of improved data analysis is accurate environmental discernment, which influences how machines see, perceive and react. “Robots are learning how to navigate the physical world and solve problems that were previously beyond their ability to learn quickly,” Witten says. With robotics, he says, we are edging closer to a technology transition like the one that occurred when the iPhone was released in 2007.

“We see the same needs for software-defined architectures in other industries,” Patel says. “And they need the same components we do in software-defined vehicles.” Each developer recognizes the need for sensor data, controls, perception management and the ability to update software. Those imperatives offer plenty of opportunities.

We are already at an ROI inflection point, Witten says. A robot can work, say, 6,000 hours in a given year, replacing a typical worker paid $15 an hour — a total of $90,000 in annual wages. Compare wage savings of $270,000 over three years to the present equipment cost of $100,000. Over time, the cost of those robots will come down, probably dramatically. And while automation adoption is largely in manufacturing today, Patel sees opportunities in hospitality industries such as restaurants and hotels as well. (The folding-laundry task may take a little longer to master, however.)

These Changes Will Begin to Affect Employment

The coming technical changes merit thoughtful, ethical contemplation of their potential effects on the workforce.

Some of those technological trends are undoubtedly good, at least for employers.

Increased automation will affect the job market. “If a robot will now do the manual lifting previously done by a human, what happens to the human?” Schroeder asks. Humans can never be wholly removed, nor should they be. Anyone implementing AI should think now about how to upskill their organization’s talent before emerging trends take off.

Ultimately, humans will always have a role — ideally, a better one. Automation can improve safety, and robotics can improve outcomes (such as helping surgeons work more efficiently), but humans are better equipped to respond to dangerous situations and to mitigate company liability.

What Is Coming Beyond 2026?

Beyond 2026, the view is fuzzier. Is that a light at the end of the tunnel, or an oncoming train? Aptiv experts have several ideas.

Communities Are Changing: An Aging Society

Due to lower birth rates and longer lifespans, the proportion of older adults in the global population is growing. That is creating shifts in economies, family structures, healthcare and transportation. “With fewer young people, fewer young people own cars. There will be more older people who are unable to drive,” Welk says. Those trends are likely to change transportation systems.

Those changes may create more demand for ride-sharing services, an increase in vehicle automation and a greater reliance on public transit and other services to transport people. This trend is already underway, but it will take years to recognize its impact. “Like most things, the changes will feel somewhat slow while we’re going through the process, but when we look back in 10 years, it will seem like the transition was very fast,” Welk says.

What Is the New Value to Deliver?

Aptiv is already working with OEMs to envision what the “next big thing” might look like. “At some point, you saturate the in-vehicle experience,” says Welk. “There’s going to be a search for new value to deliver.” He imagines that the innovations to come will build (ethically) on the data that vehicles generate. “We can potentially create new services for consumers, pedestrians, governments or other businesses,” Welk says. “Can we somehow make people’s lives easier in this connected world? Can we make their lives safer?”

For example, cellular-based sensor data may become another input that improves safety and user experiences, Patel says. In a recent Aptiv demonstration of cellular vehicle-to-everything ( C-V2X) technology, one vehicle’s sensors shared data about vulnerable road users with other nearby vehicles, using low-latency cellular network and edge compute platforms. Say a lead car perceives a bicyclist outside the line of sight of a vehicle that follows. The lead car can share the bicyclist information so the second vehicle can take appropriate action, such as slowing down – before it is even able to detect the bicyclist with its own sensors.

Another “out there” possibility is swarm intelligence, Witten says. The most visible example today is drones used as firework replacements, but expect to see additional application scenarios, in aerospace and elsewhere. They may start on a small scale, with a mix of piloted and unpiloted vehicles, where the unpiloted vehicles protect the piloted ones.

AI Results Must Be Traceable

Beyond the predictable expectation that machine learning will become faster and more powerful is the issue of making the results traceable. “This is really important for us in automotive because of safety features,” Palmisano says. “But this ‘explainability of AI’ is going to be important for enterprise applications because the liability side of it is massive.”

Computer scientists are exploring the options, such as relying on statistical results, using a wrapper to override a model and using “ ground truth” to improve accuracy.

Business Model? What Business Model?

Despite the speed of AI adoption, nobody has monetized it yet as a business case. “The only people who have monetized AI are Nvidia and the companies building data centers,” Palmisano says. Those organizations are building the infrastructure, which certainly is necessary. But the revenue side of the equation is essentially unknown. Where are the killer apps that generate revenue?

That is all part of the historical hype cycle, wherein the infrastructure is built before business models emerge. However, it also underscores where the AI-centric industry is at present, even if the timeline is being compressed.

Ideally, all of these advances will help humans improve their quality of life — and will inspire Aptiv and its partners to bring that dream to fruition.