Microwave & Terahertz Sensors For Agri-Tech Applications. Richard Dudley

Transcription

1 Microwave & Terahertz Sensors For Agri-Tech Applications Richard Dudley

2 National Physical Laboratory (NPL) The UK s national standards laboratory Founded in 1900 Located in Teddington, UK. World leading National Measurement Institute Hold base units such as meter, Kg, time, amp, etc specialists in Measurement Science State-of-the-art standards facilities The heart of the UK s National Measurement System to support business and society Government owned, to be managed by Surrey University and Strathclyde University with post-grad centre with 100+ students.

3 EM Research Free-field: Wireless Communications Capability (Antennas, EMC, standardisation) Metrology for Small Antennas & Smart Antennas Distributed/remote sensing (WSN) & Body Worn Antennas Health and Exposure (PFD, SAR) for comms, medical Terahertz Guided-Wave & comms: Guided Microwave Capability Traceable Waveform Metrology (Ultrafast) Photonic Communications (O-PCB and receivers) High Speed Electronics (PCBs) Other work: EM & other modelling capabilities Smart Electrical Grids & DC / low frequency capability Quantum Detection (electrical standards & encrypted comms)

4 EM Health/Safety & Body Worn Antennas Simplified ultra wideband phantom developed for 2.4 to 10.6 GHz Solid head torso for 0.3 to 6 GHz Liquid filled anatomic shell for 30 to 450 MHz, (also for in-body antenna testing) Balanced and low profile antennas developed for validation

5 Richard Dudley Semiconductor / High Speed Optics Physicist Research focus: Technology Communications, high speed electronics Terahertz / Far IR Spectroscopy Microwaves Applications of technology Earth observation Non-destructive testing Spectroscopy Agri-tech (field to shelf)

6 My Entry in Agri-Food-Technology I was approached by a farmer wanting to image Cauliflowers growing in the field to enable automated harvesting. Many process such as planting, weeding and feeding are already highly automated. Automated, intelligent harvesting is still limited in deployment. Crop management to ensure the maximum output from each acre requires data... Close up imaging of crops is a key enabler My challenges have been to identify imaging technology, which images through the leafy layers of a crop, identifies the internal plant composition (water, sugar, disease.) enabling robotic harvesting or sorting. Other concepts and technologies have feed from this initial challenge.

7 Todays Talk Explore my experiences with technology from ultrasound to xrays in the horticultural, agricultural, livestock and food processing environments - from seed to shelf. Provide an overview of what differing technologies can provide and limitations from a technical, deployment, safety and cost perspective. Explore some examples where I have worked or am currently working.

8 Electromagnetic Spectrum

9 Ultrasound / Acoustic Good penetration into water rich materials Propagation in air negligible Images difficult to interpret. Off shelf solutions from single point to 3D imager. Established in animal sector for quantifying hind fat thickness on pigs and cattle. Notoriously inaccurate and difficult to operate in dirty environment. Matching required between sensor head and object under test gel or gel-pack. Acoustics to monitor tree health. Connection of microphone enables stress within large trees to be assessed over time through sound. Ripeness tap test watermelons.

10 NPL Orange Seed Detection by Ultrasound Antares 3-D scanning head, clinical ultrasound. Water Bath! Positional uncertainty Seed? Skin absorption lead to false positives

11 X-Rays Excellent penetration, but ionising radiation and therefore destructive to cells and biology. 2D and 3D imagers readily available at low-cost. Has been demonstrated in-field for cauliflower imaging. Major disadvantages H&S to workers Downstream product acceptance General perception No-spectroscopy just contrast

12 Terahertz (100 GHz 10 THz) Very far-ir Instrumentation stagnated until around 1995 Thermal sources, detectors and quasi-optics Radio-astronomy main applications Short pulse laser created new THz approach Ti:Saphhire 800 nm and < 100 femtosecond pulses Non-linear crystals enabled mixing to THz frequencies with good dynamic range. Current equipment remains laboratory base, sensitive and relatively high cost. Some portable systems available but performance is reduced. Use in harsh environments currently limited.

13 Terahertz Leaf Measurements One of first images produced with THz was a leaf by AT&T Bell Labs in 1995, B.Hu, Opt. Lett. 20, 1716 (1995) Not spectroscopy, just water absorption Recent re-visits Castro-Camus, Spectroscopy.Sci. Rep.3, 2910; OI: /srep02910 (2013) Teraview corn seed

14 RF / Microwave ( 1 MHz 100 GHz) Well established measurement equipment, Costs around 1k Euro per GHz, but this is compressed at either end of frequency range above. Portable equipment available and often close in performance to lab equipment, targeted at communications and electronics applications. Measurements of dielectric constants and loss rather than absorption features.

15 Microwave Imaging Radar and microwave applications established Military applications: airport security, missile guidance... Contrast image reveals internal structure Rot, cavities, seeds, foreign bodies.. Plant structure have less water than fruit, therefore imaging through leaf layers possible. Resolution and sensitivity will be reduced. Regular observation of fruit enables development to be analysed Ripeness and disease can then be quantified. Cauliflower Images through single leaf layers 4 Lane Apple Sorting- Optical & IR

16 Microwave Imaging for Automated Harvesting Many process such as planting, weeding and feeding are already highly automated Automated, intelligent harvesting is still limited in deployment. Crop management to ensure the maximum output from each acre requires data The challenge is to identify an imaging technology, which can look beneath the leafy layers of a crop, identify the differing materials and enable precise size identification for deployment in a fully automated harvesting robot.

17 Fruit / Vegetable Microwave Measurement Surface contact critical Rough surfaces lead to error Knowledge of product under test required Absolute identification of sugar, acid etc.. Unlikely to be achieved. Good penetration into object under test ( < 5GHz) Produce Mudgett Kent Venkatesh NPL Potato ε = 65 ε =19 ε = 65 to 74 ε = 28.8 to 17 ε = 62 to 57 ε = 22 to 17 ε = 70 ε = 20 Apple ε = n/a ε =n/a ε = 42.0 ε =23.1 ε = 57 to 54 ε = 8 to 10 ε = 65 ε =19 Pear ε = n/a ε =n/a ε = 68 ε =12 ε = 67 to 64 ε = 11 to 13 ε = 71.9 ε = 10.3 Avocado ε = n/a ε =n/a ε = n/a ε =n/a ε = 47 to 45 ε =16 to 12 ε = 55 ε = 15

18 Microwave Ripeness and Sugar Content IR methods give some information on fruit ripeness, but penetration depths are questionable Microwave sensors provide dielectric spectroscopy at depth within a plant or fruit. Microwave sensors are approaching disposable cost level. NPL collaborating on three UK TSB projects to design and demonstrate a microwave Sugarcane Ripeness Sensor (Brazilian customer) Sub-surface imager for produce sorting Automated strawberry picking.

19 Black Heart in Potatoes Bench top and in-line system for identifying black heart in potatoes using UWB and contrast microwave imaging.

20 Sugar in Sugar Cane Sugarcane plantations are vast (800,000 hectors) Samples must be sent to laboratory for sugar analysis prior to harvest to effect best bio-fuel production. A hand-held or through life disposable sensor would off significant monitoring and management benefits. In partnership with Distell and Kew Gardens we are developing and testing a microwave sensor to address noncontact sugar measurement. The technology is having further interests for other produce such as sweetcorn.

21 Disease, Hydration and Ripeness The imager measures moisture and density in the plant so from the contrast in measurements you can also observe: Hydration, location, circulation and rate of water uptake in a plant. Distribution of water, uptake and unusual features measured over time can indicate disease. The swelling and water content development over time will provide information on the ripeness of the fruit.

22 Pesticide alternatives? In January 2009 twenty-two chemical pesticides were placed onto the banned substance list by the European Union despite significant lobbying from farmers. Find an alternative or face further reductions in profitability. We have investigate low-power aphid extermination methods. No heating, just gently stimulating death/implosion.

23 Earth Observation of Field Crops Satellite images are widely used to quantify crop growth using visual imagery. Satellites exist that use other types of light that can give information on plant stress and water content. For example Infra-Red, Thermal and microwave frequencies. These satellites are normally used for weather and climate predication. The quality of these images are lower than their optical counterparts. Our Proposition is: Develop a new data source by improving the image quality from weather satellite which can provide a crop hydration and stress monitoring tool. We propose combining ground based sensors with unmanned aerial vehicles (UAV) and Aircraft data with Satellite data to provide rich picture of your growing environment. The aim is to increase output from field while minimising resource use.