WebRTC will touch all of our lives over the next few years by enabling point-to-point communication, and it will also significantly impact the contact center market. I see webRTC potentially impacting customer engagement applications. Organizations will be able to engage and serve customers, patients, students and workforce through enterprise websites, mobile apps, and kiosks.
[bctt tweet=”WebRTC will touch all of our lives in the coming years by enabling point-to-point communication & customer service” username=”vyopta”]
Customer engagement applications that incorporate webRTC – like Google Helpouts – will initially be used for person-to-person interactions with limited person-to-machine interactions such as auto-attendant service for voice, video and real-time chat. Beyond point-to-point video calls, multi-party video communication scenarios tend to be even more popular. It is not traditional contact center communication nor traditional video conferencing that many have experienced in meeting rooms. There are many alternative use cases such as e-learning, customer support, and patient care that require multi-party video communication. In each of these cases, the core capability is the distribution of media streams from multiple sources to multiple destinations.
There are many different architectures that may be suitable to support these use cases, but I personally prefer the following two:
Distributive MCU
The distributive MCU architecture is a similar solution to traditional multi-party conferencing used for years with great success, but it is now much more scalable. This new technology supports communication among most endpoints, including webRTC and some legacy hardware units. This architecture allows for multi-party video and voice globally to minimize the media streams required to connect all participants. Additionally, different output streams are generated to each participating receiver endpoints, at a call quality appropriate for that device. Another advantage of a MCU solution is that it can utilize hardware decoding in the device, providing that webRTC devices include the capability to decode a single video channel in the chipset.
Distributive Router
The distributive router architecture is another solution option that achieved popularity through a proliferation of H.264 SVC infrastructure. It is the architecture being used by most of the new webRTC platforms with previously developed legacy hardware. The architecture is a hub and spoke approach, having a central point receiving a stream from every sender endpoint and sending out a stream to every participant receiver endpoint. This central point only does packet inspection and forwarding, and not processing intensive encoding and decoding of the actual media like the distributive MCU solution.
Routers provide a cheap and scalable multiparty solution, with fewer delays and little quality degradation compared with the traditional MCU solutions. However, the development and manufacture of these infrastructures are still relatively new, and adaptation of streams to different receiver endpoints – webRTC and other non-SVC endpoints – is complex. Adaptation requires support in the endpoints to generate multiple versions of the stream (i.e. with simulcast or VP8 temporal scalability) that can later be selectively forwarded in the router, depending on the capabilities of each receiver endpoint.
As webRTC evolves, as more APIs are provided, and as integration into more web browsers occurs, the vision of using video communications to transform the way organizations relate and interact with customers, patients, students, and others will become a reality.