On July 18, the Federal Communications Commission (FCC) published, in the Federal Register, its intent to allow telephone captioning providers under their Telecommunications Relay Services (TRS) to provide captions using fully Automated Speech Recognition (ASR) software. The proposed change is expected save a substantial amount of money by making the role of the human Captioning Assistants (CAs) optional.
Known as the internet Protocol Captioned Telephone Service (IP CTS), in its current form, Communication Assistants are employed to conduct the captioning. The proposed change to software generated captioning could hasten a decline in the quality of real time captioning. However, there are no current standards for accuracy or delay in telephone captioning provided under the Americans with Disabilities Act and regulated by the FCC.
Currently, captions are provided by CAs, who use software to assist in the quick delivery of accurate captions. Allowing the switch to software-generated captions is presented in the Federal Register filing as acting to “modernize and reform” captions, to save money. Making the role of CAs in providing captions optional could lead to significant cost savings.
A review of the current state of captioning is summarized here.
Switching to Automated Captions: Risking Quality Decline?
One of the challenges the FCC faces is ensuring that the switch to all-automated captions does not result in a decline in caption quality, in terms of either accuracy or timing, compared to the current CA-assisted captions. In fact, they request comment on several questions related to that problem. There is, however, no accepted method of measuring caption quality currently, and no baseline for the quality of existing, CA-assisted captions. FCC Commissioner Rosenworcel, at the FCC’s June meeting stated,
“This (proposal) is exciting. It may yield an experience for users that is comparable to older forms of IP CTS and delivers true functional equivalency under the law. But inexplicably, the FCC authorizes automatic speech recognition today but puts off for the future figuring out at what rate providers will be compensated and what service quality standards hard-of-hearing users can expect.”
Rosenworcel’s comments, which are also posted on her FCC web page, make the point that service quality standards that will allow for the comparison of automated captions to the current, CA-assisted captions, are not in place at this time. The existing criterion, functional equivalence, is a general term used in the Americans with Disabilities Act, and does not provide a basis for comparing automated captions with the current, CA-assisted captions.
The IP CTS proposal can be found here:
The public comment period for the intended switch to automated captions has begun. It ends on September 17, 2018.