Sense Media Group

Abstract:

This interview with Sara Sargent explores the rapid evolution of automotive sensing, AI, and intelligent mobility. From sensor fusion and validation to agentic in-cabin assistants and autonomous deployment, Sargent shares insights into the technologies, challenges, and industry shifts shaping the next generation of connected, context-aware vehicles.

1. AutoSens and InCabin have evolved into highly influential platforms for ADAS, AV, and in-cabin innovation. From your perspective, what are the most significant technological and strategic shifts redefining the automotive sensing ecosystem today?

Shift from independent capability development, to applications requiring a combination of interior situational understanding and exterior situational understanding – I see this first hand, from the subject matter of the technical abstracts that are submitted to our shows, to the demos in the suites at CES, understanding the value multiplier for applications with access to agents providing both sets of context is important. If people want an example of what AI looks like in automotive, consider that as the demand for contextual awareness increases, it is going to be a commodity. 

Shift from messaging around solving regulatory compliance to the validation conversation
Fairly straightforward from the title, there are many new in-cabin regulations and Automated Driving regulations, the OEMs and T1s AND suppliers need ways of validating increasingly complex systems. Therefore, companies who help solve these problems are becoming more in demand. 

Shift from basic personalization to complex agentic assistance –
Technical - you can see this in the level of complexity of the demos shown by DMS companies, mapping, audio, etc., everyone wants to be the first to provide the best agent for their domain. 

Strategic – those who win the agent race are likely to wear the crown for being the industry thought leader or expert in an industry where we’ve seen over and over that really gets you ahead.

2. As OEMs transition toward software-defined vehicles, how is the role of sensor fusion evolving across ADAS, autonomous driving, and intelligent cabin systems?

In the cabin, sensor fusion is an enabler of providing biomedical data that can be used by the system to predict whether you are having a medical incident and what it might be. Sensor fusion also makes any perception system more reliable, and robust as long as the data is used in the best way – that is where it is hard. 

There are a lot of decisions to be made when you implement sensor fusion, for every scenario you need to determine which sensors and sensor sets you will depend on if they give conflicting analysis. For instance, in a deep forest our location signal might be less reliable so we trust the camera as long as the radar and lidar match, do we want the system to have each sensor make determinations based just on that sensor data, or do we want to have an algorithm that fuses together the data first, and then that fused data is what’s ingested by the detection algorithm? 

The role is becoming really important across the board, there have already been a lot of lessons learned, and companies should have experts in sensor fusion on their teams if they decide to embark on a sensor fusion project. 

3. In-cabin technology is expanding far beyond driver monitoring into health, wellness, personalization, and immersive experiences. Which emerging applications do you believe will become mainstream within the next five years?

Agentic AI assistants and the car alerting us of ways that our habits are out of character.
 
4. AI is becoming central to both external perception systems and occupant monitoring. How do you assess the industry’s readiness to deploy AI-driven decision-making systems at scale while maintaining safety, reliability, and consumer trust?

This is a question I don’t think many want to answer and I will shamelessly duck. I think the best way is to look at how elements of the same technology or principles have been carried out in many other verticals, and start from there as a way to gauge what factors signal and allow for safe AI deployments. That’s the starting point. You also need human beings you trust to put humanity over greed or excitement for development in the position of AI decision making and influencing in your organizations and know those who you partner with for critical applications.

5. Sensor modalities such as radar, LiDAR, thermal imaging, and camera systems continue to compete and converge. How do you foresee the optimal sensor stack evolving for next-generation autonomous and semi-autonomous vehicles?

I disagree a bit, I don’t think there’s a lot of competition between these modalities, I think there are a few companies being loud about their shift from the norm. These modalities have mostly all firmly planted their use cases in the ground. You need all of these sensors for autonomous driving or you’re not doing everything you can to make the car safe, why would you turn down the opportunity to have sensor data that makes the difference between a deadly crash or not – especially when your competitors are using it? 

The evolution comes on the software side I think – what can be learned and what can be taught about driving context to the perception systems? Processing speed, new market growth/policy, and changes to system architectures are persisting influences.  

6. Cybersecurity and data privacy concerns are growing as vehicles become more connected and cabin systems collect biometric and behavioral data. How should the industry address the balance between personalization and privacy?

I’m not the right person to ask this, and actually I think this is a panel question, I’m not sure that one person should be giving an answer to this. I think an element that is important will be that the OEMs need to use transparency about privacy and data usage as an element of building brand trust and loyalty. 

7. Human-machine interaction is rapidly changing with the integration of voice AI, gesture control, adaptive displays, and contextual interfaces. What innovations in UX and HMI do you believe will define the next generation of cockpit experiences?

I think innovation will be more about all the ways the car can communicate back without language to us through haptics, light, audio, etc., things we’re learning from Human Factors researchers.  

8. The race toward higher levels of autonomy has faced both technical and commercial challenges. Are we entering a more pragmatic phase for autonomous vehicle development, and what does realistic deployment look like over the next decade?

We’re experiencing realistic deployment now, so I think it becomes more 2nd nature in the places where robotaxi deployment is active. It will become much more widespread. Those developing the map tech, sensors, and robotaxis themselves seem to have reached a level of confidence they believe is scalable – and not just them of course, but the policy makers, roadway owners, and users who allow them to operate in their jurisdictions. We’ve seen Robotaxi and autonomous trucking deployments get to where they are today from 0 on the roads a decade ago, so the next decade holds a lot of real deployments!

9. Startups continue to drive innovation in sensing, AI, and cabin technologies, yet scaling automotive solutions remains difficult. What qualities or technological capabilities distinguish the startups most likely to succeed in this ecosystem?

Survive the rigor and you must meet and make memorable impressions on as many of the right people in the industry as you can. Both the InCabin and AutoSens industries are small communities and word gets around if you are a standout whether it’s good or bad, so be kind and memorable, and be in the right place at the right time of course.

10. Looking ahead, what are the most transformative trends or breakthrough technologies you expect to emerge from the AutoSens and InCabin communities that could fundamentally reshape automotive safety and mobility?

Oh, if you want to know that then you need to come see the conference and do the demos for yourself… June 9-11 in Detroit :)