Interpreting Services at the Touch of a Buton
Interpreting technology has traditionally taken a back seat to translation automation. That situation is changing as developers focus on this next frontier in multilingual communication. This fast-growing sector is vibrant with new solutions, which move some interpreting revenue around, broaden language access, and create fresh sources of income for spoken-language service providers.
CSA Research conducted 45 interviews and demos with tech vendors, users of such technologies, and interpreting experts to get an in-depth understanding of the state of interpreting technology. In our recent research sharing the findings, we introduced the term “interpreting delivery platforms” (IDPs) to describe applications designed to support the delivery of spoken-word language services. We closely track applications for over-the-phone (OPI), remote video (VRI), remote simultaneous (RSI), and machine interpreting (MI).
OPI: Over-the-Phone Interpreting
This class describes technology that provides audio only. Users typically conference in an interpreter who assists with the back-and-forth dialogue with consecutive interpreting. For consistency with our past research, we use this term to describe audio transferred though copper wires or over digital channels. Although the term ARI (audio remote interpreting) is more encompassing, it has not been widely adopted and typically describes internet-based audio-only interpreting.
- Pros: A variety of providers offer OPI on demand across a range of languages. It is a great solution for urgent needs (healthcare, law enforcement, and emergency response teams), cost-conscious users (refugee camps, social services, and schools), contact centers where parties are already communicating by phone (airlines, banks, telecom providers, and utilities), or for businesses with unpredictable demand (such as retail outlets).
- Cons: OPI provides a poor user experience compared to other options. You are on hold a lot. The interaction is twice as long – an element in common with all consecutive interpreting scenarios, but that users we interviewed complained about more strongly on OPI calls. When the interpreter is located away from participants, the distance and lack of visual affects the dynamics of the interaction. The interpreting client may feel more isolated while the linguist is dropped into the situation with no awareness of the context.
VRI: Video Remote Interpreting
This term describes systems that manage both audio and visual delivery. To be consistent with past terminology, we use the term VRI even in cases where there is no remote interpreting with an offsite linguist. The term first appeared in the sign language industry, but is now widely used for both spoken and sign languages. This technology class saw many developments in recent years thanks to improvements in bandwidth, which previously hampered deployments.
- Pros: VRI maintains visual clues, though not necessarily completely. Being able to see participants and interpreter provides more intimacy to the meeting than in an OPI call. In scheduled situations, it is well suited for sign language interpreting and remote service delivery such as telehealth. On-demand VRI is a practical alternative when no on-site interpreter is available for last-minute requests and for brief meetings to speed up communication, such as in a public safety scenario or tourism.
- Cons: The interaction is at the mercy of the connection quality, which has historically been the reason for the slow implementation of VRI. It also requires some equipment, ranging from smartphones to dedicated devices. Hospitals are early adopters of specialized equipment and mention unit theft, sanitizing issues, and staff training as core issues. On small screens, participants have a hard time identifying with the interpreter. Linguists also mentioned the different dynamic of the interaction in the way users perceive them. Finally, participants reported that VRI doesn’t work well for large groups, in mobile situations, for long periods of time, and when dealing with a serious or emotional scenario.
RSI: Remote Simultaneous Interpreting
We introduce this new term to describe systems for the delivery of simultaneous interpreting services by phone or over the internet. The technology creates a virtual booth where interpreters – who may be in different locations – can pass the microphone back and forth as they interpret the event in real-time. RSI technology does not necessarily include a visual element such as slides or a video feed. It can handle remote speakers, remote interpreting, remote participation, and hybrid events.
- Pros: RSI maintains meeting momentum by enabling the audience to follow with just a few seconds’ lag. Companies use it to deliver webinars, workshops, seminars, group meetings, interviews, training, conference calls, live events, and hybrid on-site/remote events. They often deploy RSI at the decision-maker level for international organizations, associations, non-governmental organizations, and private corporations. This technology also provides access to individuals who rarely had access to language services before – such as businessmen traveling abroad. Interpreters report some unexpected advantages such as not having to deal with attendees coming and going in the room or not having someone standing in front of the booth and blocking the view.
- Cons: This still relatively new IDP type still needs established processes to create a smooth experience for all parties. Foreign-language participants can feel isolated. Clients often still prefer having a person physically there to give scripts to and provide instructions. Linguists coming in cold on an RSI assignment may struggle and regularly report finding this interpreting mode more demanding to deliver than in-person. They may not have a visual feed, or if they do, they may not be able to control what the camera looks at – or they may not want it because they have to focus on the already hard job of interpreting simultaneously. Some skeptics think that it is foolish to skip paying for in-person interpreting given the cost of putting together an international event.
MI: Machine Interpreting
This concept is sometimes also referred to as “spoken translation” and involves systems to process speech to text using speech recognition systems such as Nuance Dragon, then processes the text through machine translation (MT), and renders the text to speech (TTS) through speech synthesis. Applications range from out-of-the box solutions to optimized systems that you train with your acronyms, special terms, frequent misspellings, and human translation of frequent sentences. In some systems, users may get to decide whether to read or listen to the translation.
- Pros: The appeal of MI is its immediacy and low cost. Contact centers have started to use it to deflect the cost of OPI. Government agencies and the military use it in particular to resolve situations when communications are risky or nonexistent, when there is no time to get an interpreter, or when there’s too much volume of information to process otherwise. MI works great in situations where imperfect language support is better than no support at all – such as in after-hours or overflow work in contact centers. Providers target markets where organizations can’t always afford human interpreters, such as health care, press, telecommunications, and tourism.
- Cons: MI detractors cite the fact that the poor quality of the output pushes the interaction to other interpreting types anyway. The combination of the noise at each of the three steps of the process (speech recognition, machine translation, speech synthesis) compounds issues of these technology components, which are all still trying to refine their output.
When trying to identify the right system to use, many buyers and language service providers don’t understand all the parameters to analyze, nor do they know which questions to ask. They are at the mercy of vendor marketing and sales tactics – and often buy based on the “wow” factor more so than on proper system comparisons and actual testing. To learn more about how to select an IDP that is fit-for-purpose, check out our report on “Interpreters at the Push of a Button.”
About the Author