MOWing the lawn BY SATelitte

Autonomous lawnmowers are not at all like classical lawnmowers. Every day they cut a very short length of grass and leave the trimmings on the lawn as a nutritive input. They are not manually pushed, therefore saving a lot of time for the lawn’s keeper and reducing any risk of injury. They don’t use fuel so they produce the least amount of CO2 emission.

Currently available autonomous lawnmowers use underground wired perimeters for delimiting the lawn; an electromagnetic signal recognised by these machines, the strength of which is measured by a mower-embedded sensor, determines its distance from the lawn’s edge. But that wire and its subsequent installation are technical barriers for many potential customers.

In the meantime, a group of navigation satellites can help localise all the mobile objects in the world, absolutely or relatively to other known locations. So why is the famous GPS not yet in use in lawnmower robots? Or, more generally, in robotics? Is it because of the Required Navigation Performances? Price? Or both?

During the second ION Autonomous Lawnmower competition, Professor Frank Van Graas, who was accompanying the winning Ohio University’s team, told the GPS World magazine (August 2005): “the centimetre level positioning accuracy, required for lawnmowers in the contest, is actually more difficult than automatically landing an airplane”.

Description

The work is organised around a couple of market trials with early prototypes which will demonstrate the capability to resolve the lawn keepers’ problems and capture the interest of potential customers.
The main part of the project is therefore assigned to demonstrate these capabilities, with activities split into two main work packages:

• Application development: specifying functions and interfaces between the autonomous lawnmower and its guidance system, the design and development of the satellite-guidance system, adapting the autonomous lawnmower, and integrating the new guidance system on the robot;
• Trials and commercial support: preparing field trials with potential customers operating the GNSS-guided lawnmower, resulting in marketing actions and maybe sales opportunities.
  A specific R&D work package deals with the technical feasibility of ‘GNSS in robotics’, with extensive simulations regarding availability and continuity of service specific to our application, assessment of the benefits provided by the current EGNOS/EDAS and of those expected from Galileo.
• Economical feasibility: this work package aims to refine the preliminary business plan drafted in the proposal, to determine the end-users’ most important requirement and to define a marketing strategy.

Objectives

The idea sustaining this project is to increase the adoption of GNSS towards domestic robot applications, like autonomous mowers, which do not currently use GNSS-based guidance.

This would generate economic and social benefits, like reduced pollution (lawnmowers with thermal engines frequently have poor tuning), and pave the way for Galileo which, combined with GPS/EGNOS, will dramatically increase the availability of decimetre protection levels, enabling other robotic applications.

The scientific objective is to demonstrate that a local augmentation, combining a specific phase processing and the EGNOS/EDAS data, provides the GNSS receiver with an acceptable risk of non-integrity for a protection radius of a few decimetres. In the short term, that will enable the use of GNSS in domestic robotic areas where the service availability is not very stringent, as with lawn mowing.

The project also aims to demonstrate that the Galileo signals will enable many safe land activities and consequently trigger the proliferation of GNSS in several new mass market segments, and more particularly in the domain of robotics.

The implementation of the above concept in a prototype of the cheapest stand-alone GNSS technology will then illustrate the ability of a lawnmower robot to perform its mission in an efficient and autonomous way.