Speckle velocimetry for high accuracy and multi-dimensional odometry

Lead Research Organisation: Cranfield University
Department Name: Sch of Engineering

Abstract

There is great interest in improving the capabilities of autonomous land vehicles, for a diverse range of applications ranging from inspection/repair in nuclear facilities, pipeline inspections, military surveillance, search and rescue, bomb disposal/mine clearance and space exploration rovers to household vacuum cleaners, lawn mowers and pool cleaners. One area of particular interest concerns the navigation of the vehicle and in particular measuring a vehicle's movements or localisation. Odometry or 'dead reckoning' is commonly used to calculate a vehicle's position, and requires some measure of the distance travelled. Currently, the most common technique for measuring odometry involves counting wheel revolutions using wheel encoders. This is prone to errors and inaccuracies, for example due to wheel slippages, unequal wheel diameters, misalignment of wheels, surface roughness and rounding errors due to the discrete sampling of wheel increments. The research proposed here is the development of an improved method of navigation feedback using non-contact optical sensing combined with digital image processing techniques.The research proposed here is the development of an improved method of navigation feedback using non-contact optical sensing combined with digital image processing techniques. The program will involve the construction and demonstration of a test system, the optimisation of processing algorithms and an assessment of its capabilities. This will be followed by the further development of the concept to provide other navigational information about the vehicle's rotation and the detection of vehicle slippages.

Planned Impact

Beneficiaries The development of speckle velocimetry instrumentation will be of benefit to a wide range of users outside of the academic community. The main benefits of the technique will be in the improvement of existing robotics applications through access to improved navigational information and in enabling new application areas that would have otherwise been impractical. This will potentially lead to reduced operating costs, increased capabilities and increased autonomy for many applications. Realistic timescales for the benefits to be realised will depend upon the application area, however it is anticipated that benefits could be seen within 3 to 5 years of completion of the programme, as the technique will be readily integrated into existing applications. A number of potential beneficiaries in different sectors are listed below: Commercial sector: Pipeline inspections - There is a need to inspect and service underground gas and oil pipelines to detect deterioration due to corrosion and ground movement and prevent pipe leakages, cracking and deterioration. Inspection is preferable to the routine excavation and replacement of sections based upon statistical lifetime data, as it helps to reduce the cost and labour required by this maintenance work. Exploration rovers - The initial focus for this work was the application of speckle velocimetry on autonomous exploration rovers, in particular for forthcoming planetary exploration missions to Mars (EXOMARS) and the moon (MOONNEXT). The UK space industry would benefit from the research directly through our collaboration with EADS Astrium. Agriculture - There is interest in the use of autonomous vehicles in agriculture for crop spraying, land fertilisation, harvesting etc. The increased navigation information available from speckle velocimetry may allow such vehicles to move more accurately reducing the use of chemicals and fertilisers, reducing the cost and environmental impact. Public sector: Healthcare - Autonomous cleaning robots could potentially be used in hospitals, such an application would require accurate navigation so that areas are not left unclean. Mine clearance - There has been considerable interest in the use of autonomous robots for mine clearance. Such systems would require good navigational accuracy to ensure 100% coverage of the area to be swept and would reduce the need to overlap scan areas to ensure no regions are missed. Other techniques such as GPS may not be able to offer this level of accuracy and/or be unattainable. Charities involved in civilian mine clearance and the military would be potential beneficiaries of the increased capabilities of mine clearing robots. b) Communications and engagement We have an existing collaboration agreement with EADS Astrium. They are keen on further developing the technique for use on space exploration rovers and have agreed to provide support for the project in the form of the use of a rover test bed as well as help in communicating knowledge of the technique to the relevant bodies, e.g. ESA. It is intended to give presentations at international conferences in different application areas to communicate knowledge of the technique and create new links. c) Collaboration and Exploitation The proposed research includes collaboration with EADS Astrium who are interested in speckle velocimetry for application to exploration rovers. The Cranfield Business Development Office will assist on possibilities for commercial exploitation of any developments. d)Capability The PI, and where appropriate the RA's, will undertake the impact activities with the assistance of the Cranfield University Business Development Office. The PI has extensive experience with knowledge exchange. The RA's will be mentored by the PI and undertake knowledge exchange training provided by Cranfield University

Publications

10 25 50
 
Description During this research project a new method of providing odometry for autonomous exploration rovers and other robotic vehicles using laser speckle, was developed to improve upon the existing method using wheel encoders to count wheel revolutions. The technique, laser speckle velocimetry, use the motion of the laser speckle images recorded by a camera mounted on the robotic vehicle to determine the motion of the illuminated surface it is travelling over.

A simplified sensor concept using objective speckle (no imaging lens) was developed, that offers better signal levels and a simpler/cheaper sensor over subjective (imaged) speckle concepts previously studied. Laboratory tests of the technique indicate that position errors of less than 0.1% in the calculated position of the vehicle as a percentage of the distance travelled can be achieved on smooth/approximately planar surfaces such as indoor laminate flooring. In addition trials on exploration rovers have yielded errors of 0.56% (uneven sandy surface) and 1.50% (rocky ground).

Finally, the cause of the increased errors on shaped surfaces, such as those traversed by exploration rovers, was identified and investigated. With a new extended theory relating the observed speckle shift to the vehicles translation and the shape of the surface was developed theoretically and tested experimentally. This theory, has recently been published, and has the potential to allow improved speckle velocimetry measurements via new optimisations.
This work has led to two peer-reviewed journal papers, with addition publications intended once further trials with an optimised system can be performed. In addition this work proved increased research capability which will be of benefit to future projects, via training and skills achieved in the area of high speed, optimised, data acquisition and processing to meet the stringent processing limitations of exploration rovers.
Exploitation Route TThe development of the speckle velocimetry technique in this project, could potentially allow the improvement of odometry and autonomous navigation by both researchers and companies developing robots for applications that where alternative methods of positioning such as GPS are impractical or insufficiently accurate, such as pipeline inspections, exploration rovers (EADS Astrium Ltd - were project partners in this area) and mine clearance robots.

The improved theory relating speckle shift to object translation and surface shape opens new application areas for the measurement of translations over shaped objects, such as high-accuracy positioning operations in robotic manufacturing, and we intend to apply the knowledge gained during this research in this area.
Sectors Aerospace, Defence and Marine,Manufacturing, including Industrial Biotechology

 
Description There is no clearly definable economic and societal impacts from the project as yet (2014) although the RCUK, however the knowledge gained during the project enhances the research capacity of the University [RCUK - Economic and societal impact: "enhancing the research capacity, knowledge and skills of public, private and third sector organisations"] and has allow a number of follow-on applications to be submitted in the area of instrumentation for robotic manufacturing that are intended to achieve further impact such as "commercialisation and exploitation".
 
Description EPSRC Novel Instrumentation for Manufacturing
Amount £646,880 (GBP)
Funding ID EP/M020401/1 
Organisation Engineering and Physical Sciences Research Council (EPSRC) 
Sector Academic/University
Country United Kingdom
Start 06/2015 
End 05/2018
 
Description Platform Grant
Amount £1,363,268 (GBP)
Funding ID EP/N002520/1 
Organisation Engineering and Physical Sciences Research Council (EPSRC) 
Sector Academic/University
Country United Kingdom
Start 10/2015 
End 09/2020
 
Description RCUK Catapult Researchers in Residence (HVM MTC) - Application of novel optical positioning & orientation instrumentation in manufacturing robotics
Amount £50,000 (GBP)
Funding ID EP/R513519/1 
Organisation Engineering and Physical Sciences Research Council (EPSRC) 
Sector Academic/University
Country United Kingdom
Start 04/2018 
End 03/2022
 
Description Display stand at Laser World of Photonics Exhibition, Munich 2017 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Industry/Business
Results and Impact Instrumentation demonstrators from several EPSRC funded programmes exhibited at a major international exhibition & conference. The aim was to increase impact of the our research by engaging with end users. A number of useful new contacts were made that are currently being further developed.
Year(s) Of Engagement Activity 2017
URL http://openoptics.info/munich-portal/
 
Description Display stand at Photonex exhibition, UK, 2017 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Industry/Business
Results and Impact Instrumentation demonstrators from several EPSRC funded programmes exhibited at a major international exhibition & conference. The aim was to increase impact of the our research by engaging with end users. A number of useful new contacts were made that are currently being further developed.
Year(s) Of Engagement Activity 2017