Rabbit and Pork

We are Rabbit & Pork - the voice experience experts. We are a part of TIPi Group.

London, United Kingdom
Founded: 2018
More

Skills

Google Actions Development
Alexa Skills Delevopment
Voice Experience Design
Voice Search Key Phrase Tracking
Chatbot Creation

Clients

InsureAndGo
Domestic & General
LV=
CGI
Less

This promoted content is produced by a publishing partner of Open Mic. A paid-for membership product for partners of The Drum to self-publish their news, opinions and insights on thedrum.com - Find out more

Will COVID-19 accelerate the adoption of voice technology in public spaces?

by John Campbell

10 August 2020 14:22pm

At the start of the pandemic many businesses and individuals in the voice community focused on creating voice and chatbot solutions for the immediate demand. Answering questions related to covid-19, reducing reliance on call centres and providing diagnosis conversations all cropped up pretty quickly.

Now with the realisation that we are going to have to live with the virus for some time many companies are looking at ways to reduce risk for people when commuting, eating out, shopping and at the workplace.

We've seen examples of technology being implemented already, including online ordering at bars, booking a shopping slot and increased emphasis on cashless purchasing.

But what about those times when we are going to have to touch a device? How can risk be reduced? Can voice be used?

First let's look at a possible typical day highlighting those times when you come into contact with touch screen electronic devices.

  • You need to top up your Oyster card at the station. So you have to touch the ticket machine to select a top up and amount
  • Buying a morning coffee and you can pay by phone so no problems here
  • Into the office and you need to open the lift, and select your floor.
  • Lunch, you head over to the post office to send a package, the queue is huge so you use the self serve machine
  • Food time, it's Friday so over to McDonald's and you have to use the massive touch screen display to order
  • You go to the loo in the afternoon so have to push the flush button
  • Time for the gym after work to burn off the burgers. You have to type your passcode to get in and then program the running machine
  • Tesco on the way home, just a little shop so you use the self serve machine

Lots of interaction with touch screens and keypads. The best option is always going to be to wash your hands after each interaction and try not to touch your face. However let's split the above down into two types of devices and see how voice could help create a touch free experience.

1) Self serve - touch screens

  • Train ticket machine
  • Post office self serve
  • Tesco self serve
  • McDonald's ordering
  • Gym running machine

With all of above the user doesn't really have an account and all the devices are in public spaces.

Voice could be a great input option here for a couple of reasons:

  • Reduces the touch element 100% (if the user also pays with contactless card)
  • Task can be completed quicker; saying "Three Big Mac's, two medium cokes and two fries" is quicker than multiple taps
  • Discovery and task completion is easier, for when the user doesn't know the correct terminology. For example, "send the package to Norway as soon as possible" is understood as ‘first class international delivery’ by the machine in the Post Office

The only major barrier for voice technology here is how well speech-to-text performs in what can be a noisy environment and where there can be other people talking nearby.

For example in Tesco the checkout areas are noisy with the staff shouting "next customer" or noisy school kids.

However with COVID-19 procedures in place speech-to-text in these noisy environments could be easier, barriers between self serve, face masks so voices don't travel as far and people keeping distance.

We've also seen improvements in speech recognition technology to pick out one voice in noisy environments.

Are Bluetooth or mobile the key?

The other possible option is to still use voice as the input but rather than the microphones being on the self serve machine the user uses their headphone mic.

This isn't something that is too far-fetched. We already have hearing loop systems (or audio induction loops) for people with hearing aids

The problem with this is that trying to use Bluetooth to connect to a device can take time and isn’t 100% reliable. If it's not seamless people will just use the touch screen.

The other option could be control via your phone. You scan a QR code or open and app to connect and control the self serve machine.

Again not a far fetched idea, you can do this for the Alexa Firestick when you lose the physical remote and voice commands still work.

2) Simple input

  • Lift button
  • Keypad at the gym
  • Toilet flush

In these examples voice can play a role, in the lift for example to be able to say the floor you are going to, or at the keypad to say your pin.


However for some use cases light sensors will work best. Voice could be an over-engineered solution for when all you need to do is wave to flush.

So what tech will be powering voice solutions outside of the home?

In all scenarios above the solution isn't going to be an Alexa strapped to the side of a self- serve machine at Tesco with instructions of a Skill to open.

Companies will have to use voice services from companies such as IBM, Microsoft, Oracle, Nuance, Amazon and Google. All these companies offer various services, the key is putting together the different building blocks needed for your specific use case.

But what about accessibility? Voice often solves accessibility problems for those who have a physical impairment. But what about those people who don't have typical speech such as those with Cerebral Palsy, ALS, Parkinson’s Disease or Down’s Syndrome.

The majority of speech recognition software has been trained with “typical speech data” which means they struggle to understand somebody with an impediment.

However there is progress in this area. You may have already seen Google requesting training data from people with Down’s Syndrome and startup Voiceitt are working on creating speech recognition software to work with dysarthric speech.

There is still much work to be done in this area however COVID-19 will continue to accelerate the pace of voice tech being implemented in the public domain.

It will be interesting to see if we start to see such voice activated devices become common in this new normal.

Tags

voice assistant
Voice tech
voice strategy