Introduction

As a group of experienced makers from Technical University, we started this campaign for final testing and finishing the smart outlet prototype. Three of us, Electrical and Electronic Engineering students, began to develop the smart outlet last year as a bachelor thesis. Meantime, as University graduates, together with prof. Dr Andrius Usinskas we are going to finish the prototype and prepare to turn it into a mass-produced smart outlet product.

In the photo prof. Andrius Usinskas and graduate makers Tomas Kleiva, Povilas Antanavicius, and Vilius Hakas discuss PCB of the early prototype

At present the prototype has general features like monitoring and controlling lighting or electrical appliances remotely, disconnecting appliance in the case of the current increment and protecting children by switching off electricity in the scheduled time. Making a unique product, we introduced a novel feature for the prediction of the resident's habits. At first smart outlet automatically monitors habits how they switch on/off lights, turn on a washing machine or a vacuum cleaner. After that artificial intelligence algorithm in the smart outlet will start to predict well-known resident's habits. For example, the light in the bedroom will switch on automatically each working day at 6:01 a.m. and switch off 6:12 a.m.; the resident at 09:48 a.m. on Saturday will receive a reminder on his mobile phone to clean a floor. 

In the photo maker Tomas Kleiva tests the developed mobile app

We implemented all smart outlet features in the firmware for a microcontroller. The usage of tiny surface-mount devices on the two levels printed circuit boards introduce another new feature - possibility to mount a smart outlet in the outlet box. It is a significant achievement because usage of the electronics, which is installed in the outlet box is more comfortable and secure than the usage of wide available plug-in adapters as the smart outlets.

In the photo maker Povilas Antanavicius inspects soldering of the Wi-Fi module

The raised money will help us to run more advanced tests in order to evaluate the developed artificial intelligence algorithm for the prediction of the inhabitant habits. Some habits like switching of the lighting will depend from the day of the year due to the position of the Sun or inhabitants vacation time. The developed neural network will be automatically trained in the monitoring mode and will be executed in the user mode of the smart outlet. Also, the service mode of the smart outlet must be counted. 
Another part of the funds will be spent on developing the final prototype. We are going to produce an even smaller circuit board by minimising spaces between components. After running advanced tests, we will update the firmware improving prediction algorithm, fixing various bugs and optimising source code.

More information and specification is available at the dedicated website of Vilnius Gediminas Technical University smartoutlet.vgtu.lt 

Campaign video was made by Vilnius College of Technologies and Design vtdko.lt