Graduation year

2020 / 2021

Customer

Ing. Markus Starlinger

Graduates

Koblmüller_Jonas

Koblmüller Jonas
5AAT

Lang_Jakob

Lang Jakob
5AAT

Prieschl_Thomas

Prieschl Thomas
5AAT

Schlechtl_Christian

Schlechtl Christian
5AAT

Supervising tutors

Thomas Stockinger klein

Stockinger Thomas

Mitgutsch Wilfried

Mitgutsch Wilfried

Huss Arnold

Huss Arnold

Project description

In order to replace the time-consuming and labour-intensive picking of marigolds, it was the task of this project to develop a marigold harvest robot on behalf of Markus Starlinger.

This should replace human work and drive fully automatically across the field to pick marigolds at noon. The machine should be controllable with a mobile phone, which is possible through a built-in Wi-Fi router. The recognition software for these marigolds was a previous project that was programmed by three BI students. In this program, marigolds are recognized with artificial intelligence via image recognition and the exact coordinates are passed on to a PLC. With these coordinates it is possible for to approach flowers with our portal system and to collect them with our picker.

The goal for our harvesting robot was to make it as light and stable as possible, but without dispense of any functions. Batteries and a photovoltaic system were installed to insure long working hours. It was used as a roof to continuously produce electricity during the picking process. It is possible to open it in order to gain access to the switch boxes underneath. It is also possible to charge the machine with a power cable.

To drive safely across the rows in a marigold field, the machine has a strong and firm chassis. The two wide front tires with tractor profiles are driven by electric motors via a chain and ensure a steady movement of the robot. In order to prevent the machine from getting stuck, a swing axle with swivel castors is used as the rear axle, which compensates uneven floors.

A combination of a GPS-RTK system and a compass is used to keep the machine on track. This system recognizes the exact position and alignment of the robot and you can teach the machine the path over the flowers. However, this tracking is not implemented by us, but by the successor project.