AI Magazine June 2021 | Page 92

DATA ANALYTICS
personalised settings based on who ’ s using the machine . Potentially dangerous appliances like an oven could be programmed to limit features available to children for example .” As with all endeavours using personal data , however , attention must be paid to its protection . “ It ’ s not as simple as just implementing AI ,” says James . “ You need solid and reliable data to be successful , as well as having a robust data infrastructure in place . You have to look at how your algorithms work , and whether they are accurate over time . If your sensor data relates to people you must consider the ethical and moral use of that data , and how your algorithms guide choices that impact upon people ’ s lives .”
With the two technologies operating in unison , future possibilities are potentially transformative , as Lippett explains : “ For example , the next generation of smart speakers could support connected healthcare , monitoring biometric data — heart rate , breathing rate , temperature – to predict and prevent an illness before the user is even aware . Smart cities will be capable of delivering you to the only empty parking space in a busy city , redirect traffic to minimise congestion , or conserve energy by turning off streetlights that aren ’ t needed .” The ongoing COVID-19 pandemic has played its part too in accelerating the technology ’ s development . “ The last 12 months has seen the world adopt remote working as the norm , so anything that reduces direct or human interaction is now at the top of corporate and personal agendas ,” says James . “ In addition there are huge future

“ We need to be able to minimise the reliance of smart devices on the cloud and begin transitioning towards an IoT model that integrates AI locally ”

MARK LIPPETT CEO , XMOS
92 June 2021