[Blog] Always-On Vehicle Monitoring—Without the Battery Drain

Drivers expect their vehicles to be safe whether they are behind the wheel or not. But delivering on that expectation, with a system that monitors, detects, and responds around the clock, has proven harder than it sounds.

Last year saw car thefts in the US drop to their lowest level in decades, down 23% year over year. Still, a theft occurs every 48 seconds, causing staggering losses for individuals and organizations. Always-on monitoring has emerged as a potential solution, enabling smarter alarms, evidence gathering, and advanced security measures. The challenge is how to keep such a system powered around the clock without draining the vehicle’s battery in the process.

In a recent webinar, Lattice explained how its FPGAs and External Monitoring System (EMS) architecture can help designers address this challenge and advance the automotive industry.

The Always-On Problem
In theory, comprehensive EMS has numerous benefits, from safety and security to personalization and convenience. In practice, however, traditional EMS designs deliver on those promises at the expense of something more fundamental: reliability.

Vehicle monitoring systems rely on built-in central compute units that consume up to 300 watts when fully operational. Run continuously, that adds up to about 10% of a typical electric vehicle (EV)’s battery capacity. Low power and domain controller modes can lower this to 15-60 watts, but that is still too much for most 12V systems to sustain. Even aftermarket dash cams, at up to 44 watts, can drain a standard battery within days.

In short, standard automotive power infrastructure is not up to the task. If a car owner takes a few weeks off from driving, they risk returning to a dead battery. That’s not a trade-off most consumers are willing to make.

Event-Driven Intelligence at the Edge: Built on FPGAs
Lattice’s approach positions Lattice CrossLink^™-NX FPGAs directly in the data path, sitting between camera sensors and their serializers, enabling event-driven decisions without waking the full system until a real threat is detected.

Each FPGA acts as a real-time triage layer, evaluating three questions before waking the rest of the system:

1. Is there motion in the camera’s view?
2. Is a human present?
3. Is the person actually interacting with the vehicle?

Only if all three are “yes” does the system activate and route data to the central compute unit for analysis.

CrossLink-NX FPGAs are purpose-built for embedded vision in resource-constrained builds, offering:

- Low power functionality. Built on the Lattice Nexus^™ FPGA Platform, CrossLink-NX FPGAs use up to 75% less power during operations than similar options.
- A compact footprint. The CrossLink-NX family includes FPGAs in a range of sizes, with the smallest model measuring just 4 mm x 4 mm.
- Dynamic scaling. CrossLink-NX FPGAs use on-chip logic to adapt in real time; the system does only what’s necessary to maintain operations, reducing consumption.
- Reprogrammability. FPGAs can be reprogrammed as needs change, extending the life of in-vehicle systems.
- Interoperability. This solution integrates cleanly into existing in-vehicle workflows, enabling deployment without affecting existing advanced driver assistance systems or driving functions.

This approach brings total system power consumption down to approximately 200 milliwatts on a standard four-camera system, translating to roughly 75 days of continuous monitoring on a 12V battery.

Beyond Security: EMS as an Accelerator
That same infrastructure also enables personalization and access control features:

1. Driver recognition
   The system identifies approaching drivers and pre-adjusts seat, climate, and media preferences before they reach the door, with potential for biometric keyless entry.
2. Zone-specific access
   The system controls door access based on approach angle: opening only the trunk for rear approaches, or granting passenger-door access to registered rideshare riders.
3. Time-limited and role-based access
   Rental car providers, transportation fleet managers, or even families may appreciate time- or role-based biometric access, allowing the primary owner to control when, where, or how a vehicle can be operated by a given driver.

These are just a few examples of what Lattice’s EMS can make possible, with many more possibilities ahead. Automakers who adopt this architecture can differentiate driver experiences and bring next-generation features to market first.

Making Always-On a Reality
For years, always-on exterior monitoring has been a feature drivers want and carmakers want to deliver, but traditional architectures could not support it without significant trade-offs. Lattice’s FPGA-based approach changes that, delivering the security, personalization, and access control features the market has been waiting for, within the power budget modern vehicles actually have.

To dive deeper into Lattice’s approach to EMS architecture, watch the full webinar or download our solution brief on always-on vehicle video security. To discuss how our technology can help build the next generation of in-vehicle automation, contact our team today.