02 September 2019 | Andreas Koenig
We need to be more realistic about automation, explains Andreas Koenig.
AI and automation's evolution is exciting, but companies seeking rapid improvements in productivity see the ROI model fails to add up. From the time and cost associated with automation to the constraints of simplistic application, full automation isn't always an option.
Consider warehouse efficiency: it has never been more important, but continued labour shortages mean that it has never been more challenging. With the need to improve productivity, businesses are intrigued by the promise of automation. The concept of lights-out operations that can run 24/7, 365 days a year is compelling, but the reality is different.
Full automation is not achievable today. Firstly, it is complex. There is no switch for a completely automated environment. It takes time and money to train machines to perform tasks - and few organisations have the skills to create the deep learning structure required.
Given the speed of innovation within many product cycles, time is a factor. If an item is only in production for a few weeks, there is not enough time to train the technology to automate the process or achieve an ROI.
While progress within this field is impressive, automation technology can currently manage simple processes only. Technology vendors will confirm: robots cannot handle processes with multiple intricate steps and machines cannot comprehend events outside their sphere of reference.
There are opportunities to augment the performance of the existing human workforce through wearable devices. Workers will be able to leverage an ecosystem of wearables, such as gloves with detachable barcode scanners, to enhance human collaboration and human-to-machine interaction. From gloves to glasses and cameras, the augmented human workforce will achieve more productivity in a safer and, hopefully, happier fashion.
Andreas Koenig is CEO at ProGlove