Against this backdrop, intelligent cockpits have become the focus of intelligent competition among automakers, and how to create a more ideal human-computer interaction has become one of the key factors affecting the elimination process.
In summary, multimodal interaction integration is a development trend, and future interactions will incorporate perceptions such as mood and body sensation, thereby enriching and deepening the intelligent cockpit industry chain. Ultimately, achieving a truly intelligent human-computer interaction experience requires the efforts and cooperation of every link in the upstream and downstream of the industry chain.
Proactive multimodal interaction is the future development trend.
After years of development, in-vehicle human-machine interaction now mostly combines various methods such as buttons, touch, voice, and biometrics, and is integrated into multiple areas such as in-vehicle display systems, smart seats, and interiors to create intelligent and personalized cabin spaces.
Admittedly, intelligent interaction modes are becoming increasingly diverse, but what's more important is how to establish the integration and connection between these modes. "Simply adding more interaction modes will result in limited consumer satisfaction," an employee of a parts company commented from a developmental perspective.
This, to a certain extent, determines that one of the mainstream trends in smart cockpits is towards multimodal fusion and interaction. In the product plans of major automakers from 2020 to 2023, we can see that multimodal interaction in smart cockpits has become a key product focus.
Compared to simply enriching the interaction modes of smart cockpits, the industry believes that multimodal interaction can achieve rapid and accurate understanding and fulfillment of user needs. For example, when the voice input is "I want to see the food street behind this building," the user needs to say a complete sentence to achieve the interaction goal. Moreover, the words that users frequently use may have multiple meanings or ambiguities, such as "this building" and "that street." However, if the user adopts a multimodal interaction method, the combination of voice and eye tracking can quickly complete the interaction goal.
Another trend is the development of functions based on usage scenarios, with smart cockpits gradually providing proactive interactive services. Taking the Changan UNI-T as an example, its AI camera located near the rearview mirror can perceive the in-car environment in real time, identify the user's contextual needs, and proactively provide corresponding services. If it detects that a person in the front seat is smoking, the system will automatically activate the smoking mode and partially open the windows for ventilation.
In summary, the current stage of intelligent cockpit development is one of capability growth, while human-machine interaction is still in its early stages. Previous intelligent cockpits focused more on voice control; to improve the overall intelligence of the cockpit, more sensors are needed to perceive the in-vehicle status.
Therefore, Gasgoo Automotive Research Institute concludes that: on the one hand, voice-centric multimodal interaction is one of the future development directions, moving from "passively hearing needs" to "fulfilling needs"—a precise but passive process. On the other hand, vision-centric proactive interaction is also one of the future development directions, moving from "actively discovering needs" to "fulfilling needs"—a process that is ambiguous but proactive. Therefore, the combination of voice and vision to form proactive multimodal interaction is the future development trend.
How to achieve a more ideal human-computer interaction?
Chinese consumers also have a high willingness to pay for digital cockpit experiences. According to the "Survey on Chinese Consumers' Use of Car Sharing," nearly 50% of consumers expressed strong interest and willingness to pay for cars with intelligent sensing and interaction capabilities.
Admittedly, there are currently a variety of in-vehicle interaction methods and a large market for them, but to achieve ideal human-computer interaction, breakthroughs are still needed in terms of technology and logical design.
In practical applications, some existing interaction methods have not improved the user experience. An insider at NIO told Gasgoo, "Some interaction methods exist more to prove the feasibility of the technology; they look cool, but are actually useless."
Before incorporating interactive features, manufacturers need to conduct thorough research and deeply understand user needs, rather than simply indulging in self-indulgence, the aforementioned insider added.
Of course, understanding user needs is only the first step. For manufacturers, technological feasibility also needs to be considered. For example, compared to single-mode interaction, multimodal interaction places higher demands on algorithms and chip computing power. However, with the support of high-performance chips, a fully intelligent cockpit may be realized in the near future.
This means that to achieve ideal human-computer interaction, at least two conditions need to be met: firstly, it is necessary to improve multimodal perception technologies, including voice and gestures; secondly, it is necessary to build a new human-vehicle ecosystem around user needs.
Ultimately, from a manufacturer's perspective, to stand out in a highly competitive market, they must focus on building core competitiveness through logical design, technology, and data. Regarding product differentiation, manufacturers can analyze user needs, uncover those needs throughout the entire vehicle ownership process, and customize corresponding strategies.
In light of this, some in the industry believe that OEMs could potentially partner with technology companies and traditional suppliers to maximize their supply chain advantages. On one hand, OEMs have a better understanding of customer needs and can control R&D and production schedules, possessing the foundation for integration; on the other hand, technology companies have a greater advantage in the technological field, while suppliers can provide more flexible capabilities, facilitating coupling and control by OEMs.
