Supervision of Banana Transports by the Intelligent Container - - PowerPoint PPT Presentation

Supervision of Banana Transports by the Intelligent Container

Coldchain Manangement. 4th International Workshop, 27./28. September 2010, Bonn Reiner Jedermann, University Bremen, IMSAS / MCB Axel Moehrke, Dole Europe Import BVBA, Belgium

Autonomous logistics SFB 637

MCB MCB

2

The banana challenge

• Introduction (by Axel Moehrke)
• Bananas can produce up to 800 Watt of

heat per ton during ripening

• Hard to control after a certain point
• Improve processes along the transport and supply

chain by permanent monitoring of quality changes

• Untimely ripening is the greatest risk
• Can be triggered in the field, in the packing plant or

during transport

• Or by insufficient cooling / temperature changes

Full and accurate temperature supervision and

control is a precondition for improvements in the banana chain.

Autonomous logistics SFB 637

MCB MCB

3

Online monitoring by the intelligent container

• Real-Time remote monitoring of local temperature

deviations

• Idea first presented 2006
• Transfer project 2008 + 2009
• Online Access
• Focus on results from field tests
• Outlook to future research

Sensor Nodes External Communication Gateway Pre-Processing

Autonomous logistics SFB 637

MCB MCB

4

Outline

• Set up for field test
• Observed temperature deviations
• Over length of container
• Between different containers
• Inside one box
• Required sensor density
• Intelligent data processing
• The future of the intelligent container

Autonomous logistics SFB 637

MCB MCB

5

Installation in Costa Rica

Gateway

Autonomous logistics SFB 637

MCB MCB

6

Installation in Costa Rica

• 4 pallets equipped with 4

sensors each

S S S S S

Tier 1 Tier 8 Air ducts Foam Block

Autonomous logistics SFB 637

MCB MCB

7

External Communication

• Forwarding data from sensors to web-server
• Using the vessel’s email system

Web- Server WLAN Satellite Wireless Sensors Gateway Internet Email Server Ship Container

Autonomous logistics SFB 637

MCB MCB

8

Webinterface

Mote-ID Last message: Temp. Humidity Voltage 1 2009-09-23 14:00:02 14.79 °C 93.0 % 2.8 V 2 2009-09-23 14:00:02 13.96 °C 94.0 % 2.81 V 3 2009-09-23 14:00:02 14.34 °C 95.0 % 2.84 V 4 2009-09-23 14:00:02 13.69 °C 81.0 % 2.8 V 5 2009-09-23 14:00:02 14.36 °C 102.0% 2.86 V 6 2009-09-23 14:00:02 15.2 °C 82.0 % 2.86 V 7 2009-09-23 14:00:02 15.57 °C 100.0% 2.75 V 8 2009-09-22 02:00:02 15.3 °C 97.0 % 2.82 V

Autonomous logistics SFB 637

MCB MCB

9

Signal attenuation by the fruits Distance 0.5 meter

• ⅓ of all links

completely failed

• ⅓ of all links was not

available part of the time

• ⅓ of all links worked

well most of the time

• Partly compensated

by network

• New hardware?

2 4 6 8 10 12 0.5 1 1.5 2 2.5 Pallet

19

Pallet

14

Pallet

8

Pallet

1

Door End Refrigeration Unit G a t e w a y

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Distance to cooling unit in [m] Distance to floor in [m] Packet Rate: >0.75 0.33< <0.75 <0.33

• Water-containing goods hinder the radio communication of wireless

sensors @ 2.4 GHz

Autonomous logistics SFB 637

MCB MCB

10

The necessity for core temperature measurement

• The internal sensors of the aggregate help only little to

estimate the banana temperature

4 6 8 10 12 14 16 18 20 14 16 18 20 22 24 26 28

Days in August 2010 Temperature in [°C]

Pallet core reefer side Pallet core door side Pallet 11 mid Air supply (Aggregate sensor) Air return (Aggregate sensor)

Autonomous logistics SFB 637

MCB MCB

11

Temperature difference over length of container

• Bananas are cooled down by the container
• Large differences in time required to achieve 17 °C

10 12 14 16 18 20 22 24 14 16 18 20 22 24 26 28

Time in [days] in September 2009 Pallet Core Temperature in [°C]

Reefer Side Container 1 Door Side Container 1 Reefer Side Container 2 Door Side Container 2 Air Supply

Reefer side 2.4 days Door side 6.35 days

• Door-end 

Reefer end

• Two containers of

same type and year of manufacture

Autonomous logistics SFB 637

MCB MCB

12

Comparison of 4 Experiments

• Large variations in age of containers
• Newer equipment cools faster, but local variations

cannot be avoided

• The hot-spot can be at the door end or somewhere in

the middle of the container

2 4 6 8 10 12 2008 2009(A) 2009(B) 2010

Age 9 years Age 12 years Age 12 years Age 2 years

Time to cool down to 17 °C in [days]

Reefer End Door End Maximum

Autonomous logistics SFB 637

MCB MCB

13

Location of hot-spots

• Horizontal cut through the container
• Average pallet core temperature in tier 5 (1.25 meter above floor)
• Pallet 11 in the middle / left side is the hottest, but neighbor pallet
• n the right side almost normal.

2 4 6 8 10 12 15.4 15.6 15.8 16 16.2 16.4 16.6

Pallet 1 Pallet 1 Pallet 3 Pallet 11 Pallet 11 Pallet 10 Pallet 19

16.20°C 15.80°C 16.42°C 16.59°C 16.53°C 16.21°C 16.11°C August 2010

Distance to cooling unit in [m] Average temperature in [°C]

Hot-spot Left side Right side

Door

Air supply

Autonomous logistics SFB 637

MCB MCB

14

Required sensor density

• 4 Sensors are not sufficient
• Pallet at aggregate and door side / lowest and highest tier
• But where to place

the extra sensors?

10 12 14 16 18 20 22 24 14 16 18 20 22 24 26 28

Container BH

Time in Days in September 2009 Temperature [°C]

Full Temperature Range Range covered by corner sensors Corners Sensors: [ 5 8 17 19 ] Extra Sensors: [ 7 13 ]

2 4 6 8 10 12 0.5 1 1.5 2 2.5 Pallet

19

Pallet

14

Pallet

8

Pallet

1

Door End Cooling Unit Base

16.0 17.8 18.5 17.9 16.7 16.8 16.7 16.9 15.8 16.3 16.5

• 15.9

16.2 16.1

• Distance to cooling unit in [m]

Distance to floor in [m] Average Box Temperature in [°C]

Autonomous logistics SFB 637

MCB MCB

15

4 6 8 10 12 14 16 18 20 14 16 18 20 22 24 26 28

Days in August 2010 Temperature in [°C]

Accuracy of temperature measurements

• How accurate can we measure temperature inside a

banana box?

• Cooling air flows through the boxes  variations
• Even loggers in similar positions (distance ~ 5 cm)

≈ 0.1 °C at end of transport Up to 1 °C during cool down

Temperature difference in

• ne Box

Autonomous logistics SFB 637

MCB MCB

16

Accuracy of temperature measurements

• Pulp temperature measurement
• Hurts the banana
• Comparison: Logger temperature before opening of pallets 

Pulp temperature  Logger in average 0.05 °C too high (σ = 0.085 °C)

• Pulp temperature in centre of box 0.2 °C … 0.4 °C higher than

side of box

Autonomous logistics SFB 637

MCB MCB

17

Innovation Alliance

• New project starts in September 2010
• 13 partner companies and 6 research institutes

Sea-Containers (Bananas) Truck-Transports (Meat) Software RFID and Electronics

Federal Ministry of Education and Research

Ethylene Sensor

Autonomous logistics SFB 637

MCB MCB

18

Intelligent data processing

• Not only forward data, but pre-

process

• Different algorithms (OSGi

Software-Bundles) on demand, similar to App-Store for mobile phones

• Elements of the decision

support tool

• Spatial interpolation of

temperature by Kriging

• Prediction of future

temperature curve

• Shelf life models

High sensor tolerance

Autonomous logistics SFB 637

MCB MCB

19

10 15 20 25 30 35 14.5 15 15.5 16 16.5 17 15.14°C 15.83°C 16.07°C 15.80°C 16.05°C 16.27°C 16.20°C 15.73°C 16.15°C 16.06°C 16.21°C 16.50°C 16.56°C 15.42°C 15.56°C 16.57°C 16.53°C 16.85°C 16.35°C 16.59°C 15.68°C 15.87°C 16.38°C 16.11°C 15.82°C 15.84°C 16.42°C

Crown Rot Rotten Finger Rotten Finger Button Number Temperature in [°C]

Relation temperature and quality

• No relation between box temperature and defects per box
• Further influence factors have to be checked
• O2, CO2, Ethylene
• Age at harvest
• Micro biological load
• Mechanical damages

Autonomous logistics SFB 637

MCB MCB

20

0.2 0.4 0.6 0.8 1 2008 2009(A) 2009(B) 2010

Average defective fingers per box

Rotten Fingers Crown Rot Anthracnose

Relation temperature and quality

• Relation to average quality per container
• But no statistical evidence
• Only on container level
• No relation inside one container

2 4 6 8 10 12 2008 2009(A) 2009(B) 2010

Time to cool down to 17 °C in [days]

Reefer End Door End Maximum

No data available

Autonomous logistics SFB 637

MCB MCB

21

Summary

• The pallet core temperature can show large variations
• Differences in cool-down time
• Temperature is tricky to measure
• The hot-spot can be anywhere, will be missed if only 4 sensors
• Even inside one banana box ΔT ≈ 0.5 °C
• Relation of quality and temperature needs further

evaluation

• Other factors likes gases and biological variance
• The system will help to reduce losses by unwanted

ripening

• Accurate temperature monitoring is the basis for the further steps
• Compensate different quality levels by FEFO planning

Autonomous logistics SFB 637

MCB MCB

22

The End

Thanks for your attention www.intelligentcontainer.com

• Dr. Ing. Reiner Jedermann

University Bremen, FB1

Institute for Microsensors, -actors and –systems (IMSAS) Otto-Hahn-Allee, NW1 D-28359 Bremen, GERMANY Phone +49 421 218 62603, Fax +49 421 218 98 62603 Email rjedermann@imsas.uni-bremen.de Jedermann, R: Autonome Sensorsysteme in der Transport- und Lebensmittellogistik, Dissertation Universität Bremen, Verlag Dr. Hut, 2009.

Autonomous logistics SFB 637

MCB MCB

23

Effects of interruptions of the power supply

• Interruption of power supply during transshipment

(Harbor Costa Rica)

• Only small rise

inside pallet < 0.2 °C per hour

1d 1d6h 1d12h 1d18h 2d 2d6h 2d12h 2d18h 3d 14 15 16 17 18 19 20 21 22 23 24 25

Time in [Days/Hours] in May 2008 Temperature in [°C]

Pallet Core Pallet Surface Wall Return Air Grid Air Ducts Floor (Supply)