PULP-SEED: Philippines UK coLlaborative Partnership-System for Environmental and Efficient Drying

PULP-SEED is a British Council Newton Fund Institutional Link between Coventry University, UK and the University of San Carlos, Philippines.

The institutional link focuses on three areas:
  • Technological advances in factory monitoring systems to enable creation of new process models and support follow-on automation;
  • Developmental activities to result in staff training for research skills at USC and lessons to be applied to CU
  • Scoping industry-academia links and link-forming know-how at USC
All three areas will lead to social and economic impact either directly (factory workplaces and promotion of economic model to other factories) or indirectly (empowering the Filipino academics to create wealth and social benefit through high standard research).

This website presents the PULP-SEED project and provides access to 'live' data from a Wireless Sensor Network (WSN), deployed in a mango waste processing factory in Cebu, Phillipines. This WSN is a key component of the first focus area 'technological advances in factory monitoring systems to enable creation of new process models and support follow-on automation'

PULP-SEED: Philippines UK coLlaborative Partnership-System for Environmental and Efficient Drying

The Philippines produces around 1 million tonnes of mangoes per year, with more than half of this quantity being processed into products such as juices, dried fruits, fruit bars, candies and jellies. The mango processing industries use only about half of the mango fruit (by mass), with the remainder being waste such as the peels, seed husk and seed kernel. This mango waste is simply disposed of in open dumpsites and left to rot, releasing foul odours and potentially hazardous leachate. The waste is also a health hazard, as people scavenging dumpsites often eat the freshly discarded mango parts.

Recently, a start-up company, GEM, was formed out of the University of San Carlos (USC) Department of Chemical Engineering that converts waste from the mango processing industry into useful products, using novel technology developed at USC. The company began operations in 2012 with initial seed funding of $2 million. The company built its 2,500 m^2 facility in a one-hectare land area in Bangkal, Lapu-lapu City, Cebu.

The aim of this focus area is to make use of Wireless Sensor Network (WSN) technology to improve the yield and efficiency of existing mango waste processing. The outcome will therefore be to increase the viability of mango waste processing such that it can be scaled up to process all or most of the available mango waste currently being produced (current processing only takes waste from 1 of 36 plants in the region). This endeavor will lead to job creation and increased social wealth in disadvantaged communities in the local region. Furthermore, it is envisaged that another positive outcome of this work will be to inspire other Filipino industries to make greater use of their academic institutions as a source of technological innovation.

  1. Solar drying facility where direct solar energy is harnessed and used as an alternative power source for drying of raw mango waste peels and seeds. WSNs can be used to monitor the environmental and process conditions inside the facility such as temperature, humidity, light intensity and so on, in order to characterise and model its drying mechanisms and properties. The data and information gathered by the use of WSNs can help to further improve the design of the dryer. The solar power generation system can then achieve high productivity, yield and cost - efficiency. Further, WSNs can assist in studying the effects of direct solar heat on the characteristics and properties of dried products.
  2. WSNs will be deployed in a prototype tunnel dryer used in the facility, which is used to dry the high - value products from mango wastes, such as mango flour, mango tea, pectin, and others. Data obtained with the help of WSNs will be used to study the effects of drying conditions and parameters in the product formulation and to describe drying mechanisms. Optimisation of process conditions and operating parameters can be achieved with the use of WSNs.

The basis for optimisation of processing will be through forming statistical models of the processes that incorporate empirical data. These models will then be used to optimise the process in terms of efficiency and quality of production through automation techniques.

This project will have impact in the following areas:
  • GEM will benefit from more efficient production processes and logistics based on the research collaboration outputs.
  • The plant's current workforces are drawn from the local community. Therefore, enhancing the viability and profitability of the mango waste processing plant will provide a boost to the local community by enabling a larger work force.
  • With the help of CU, USC will build a strong research and skill base in wireless sensing technology.

To meet the aims of the project the following objectives have been set:
  1. Analysis of the current monitoring process in place at GEM and the current situation including wastage.
  2. Analysis of the monitoring requirements to diagnose quality and yield of mango waste processing.
  3. Development of a prototype system for deployment in a mango solar dryer.
  4. Development of an optimal drying strategy using both data collected from the WSN and a record of yields and quality of the resulting produce.
  5. Development of automation techniques to manage the solar dryer environment to maximise yield and quality.

PULP-SEED: The Factory

Green Enviro Management Systems, Inc. (GEMS) (http://www.greenenviromanagementsystems.com/) is engaged in holistic and integrated approaches to (food & agricultural) waste management.

GEMS have dveloped two mango waste drying processes in the factory i) a solar dryer and ii) a tunnel dryer. Both drying methods process three main mango by-products---mango seeds, mango kernels and mango husks. The process fully converts this raw material into useful products, leaving no waste at all. Two drying processes namely solar and tunnel drying are presented here.

The solar dryer initiates drying of wet mango waste. Once dry the wastge is grounded using a screw press. The resultent material is then further dried in the tunnel dryer.

The Solar Dryer

The solar dryer is a brick building with a transparent polycarbonate roof measuring 30m X 30m X 3m. Within the solar dryer there are 36 drying racks (5.55m X 1.15m X 2m) with 5 drying trays. The flooring of each tray is made from nylon netting to allow sunlight the penetrate lower levels.

In the solar dryer wet mango waste is spread on the trays twice a day at 8am and 4pm. This can not be done at any other time due to extreme heat and risk to workers. The mango placed at 8am is usually fully dried by 4pm when the new batch is laid out. There is little drying during night time but the mango peels and seeds placed at 4pm will be ready for the following day.

The Tunnel Dryer

The tunnel dryer measures 11m X 1.5m X 1.5m and is made entirely of stainless steel. The tunnel has ramps on both sides to allow the racks to be loaded/unloaded into the dryer. On one side of the tunnel are ten 1kW resistive heaters connected to a temperature control panel on the outside. The other side of the tunnel has 5 circulation fans, with 5 exhaust fans directly above to remove cold air.

To dry the mango waste, nine racks of the waste are loaded into the tunnel dryer and both ends of the tunnel are sealed. Once ready the dryer is switched on for four hours with a set point temperature of 65 degrees Celsius. After 4 hours the dried product is unloaded and the workers check if the mango waste has dried. If not, the product is stirred on its tray to remix so wet layers are exposed for the next drying time. The mango waste will then undergo another set of drying for four hours until completely dry.

PULP-SEED: Wireless Sensor Network

The PULP system is a full end-to-end WSN environmental monitoring system targeted at collection of data from the drying environment. PULP is designed to gather sensor data (temperature, humidity, radiant heat, air flow) and transmit that data to a central database where it can be viewed with a web-browser (this site).

The diagram above shows the an overview of the end-to-end system architecture. Data flows from the sensor nodes to a gateway, where it is transmitted to a remote server via the Internet and made available to user applications, such as a data analysis portal and further used to create a mango drying model. To reduce risk and development time, we opted to use off-the-shelf hardware and software wherever possible.

System Hardware Description

Our sensor node combines the TelosB platform with a custom interface board. The TelosB is based around an MSP430 F1611 CPU and a 2.4 GHz CC2420 802.15.4 radio. The TelosB also includes an integrated temperature and relative humidity sensor---the Sensirion SHT11. Our custom interface board provides input for one black-bulb sensor and one air flow sensor. Each sensor node is packaged in an off-the-shelf IP65 plastic enclosure with an slot for the air flow sensor. Nodes are affixed to the shelf supports using two cable ties.

The gateway was built using a Raspberry Pi 2 model B combined with a TelosB node.

System Software Description

The diagram below shows the system flow. A set of WSN nodes transmit their data. The sink node (connected to the gateway) receives packets from each individual sensor node and forwards these messages via USB to the gateway server. The gateway server receives these messages through the SerialForwarder program, which opens a packet source (in this case the USB-connected sink node) and lets many applications connect to it using TCP/IP streams in order to use that source. FlatLogger is one such application. FlatLogger receives packets from SerialForwarder, extracts the relevant data, and logs the data to a flat file. Each hour a Push Data script is run which transmits the logged data to a PHP page on the remote server where it is archived. Finally this data can be interrogated and downloaded via a data portal hosted on the remote server.

This system is open source and the code can be found at https://github.com/jbrusey/cogent-house/tree/pulp.

PULP-SEED: Deployment Description

From August 2015 our system has been deployed in the GEM factory in Cebu, Philippines. The main goals of the deployment are to: i) validate the performance of our system in-situ, and ii) collect an archive of data for offline analysis which can be used to build the drying model.

The PULP system has been deployed in 2 areas of the GEM factory:

Solar Dryer Deployment

The solar dryer is the factories primary method of drying out mango peel and kernels. A total of 30 nodes have been deployed in the solar dryer transmitting to a single server (Pulp1). 15 nodes (Nodes 1--15) have been deployed on a bracket at 2.13m in height. The other 15 nodes (Nodes 15--30) have been deployed on a bracket lower down at 0.6m. The nodes have been deployed in a 5m X 5m grid.

Tunnel Dryer Deployment

Within the drying tunnel a total of 10 nodes (Nodes 31--40) have been deployed. Since the tunnel is made from stainless steel, the server (Pulp2) is placed outside of the casing, with the sink node passed through a hole cut out into the tunnel.

PULP-SEED: Project Team

The team working on this project is a multi-disciplinary group consisting of electrical engineers, computer scientists, and chemical engineers from both Coventry University and the University of San Carlos. The following table lists the members of this project.

to

Instructions

To display data for a particular date:
  1. Select the corresponding Date range.
  2. Select the type of sensor by clicking the Select sensor option.
  3. Specify the nodes you want the data from. (Refer to Deployment Map)
  4. You may download the plot by clicking on the Download Plot button.
to

Instructions

To display data for a particular date:
  1. Select the corresponding Date range.
  2. Select the type of sensor by clicking the Select sensor option.
  3. Specify the nodes you want the data from. (Refer to Deployment Map)
  4. You may download the plot by clicking on the Download Plot button.
Download Plot

Data Yield for PULP1 (Green house) and PULP2 (Tunnel dryer) Sensor Deployments

to

Instructions

This plot shows the data yield of the deployed sensors for PULP 1 (greenhouse dryer) and PULP 2 (tunnel dryer). It shows the individual node yield over time. Please refer to Deployment Map to find location of a particular node in the system. To display data:
  1. Select the corresponding Date range.
  2. Select which server(s) you'd like to view.

Plot for the number of data pushes from the remote to the central server.

Download Plot
to

Instructions

This plot shows server status for PULP 1 (greenhouse dryer) and PULP 2 (tunnel dryer). We expect 24 pushes per day (one data push per hour). To display data:
  1. Select the corresponding Date range.
  2. Select which server(s) you'd like to view.

PULP-SEED Deployment Log

The deployment log indicates the dates when PULP1 and PULP2 were deployed. It also indicates the changes done on each deployment.

PULP-SEED Data Download Page

to
Download

Instructions

This page lets you download data from the PULP-SEED Project in CSV format. You may specify the inclusive dates, and nodes from the contraols above. The downloaded file has 11 columns:

  • Time-the timestamp when the data was collected (unix timestamp)
  • NodeId - the node name as specified in the deployment map for PULP1 and PULP2.
  • Temperature-the air temperature in degrees Celsius.
  • Humidity - the % relative humidity.
  • Solar - the solar sensor analog voltage value.
  • AirFlow - the airflow analog voltage value.
  • Blackbulb - the blackbulb analog voltage value.
  • Voltage - this is the battery voltage a sensor node has.

PULP-SEED: Project Document Repository

All public documents are listed below

Technical Reports

Title: PULP-SEED Factory WSN Deployment
Author(s): Ross Wilkins, Elena Gaura, James Brusey, John Kemp
Publication date: January 2016
Abstract:
This report presents initial monitoring results, forming part of the focus area: 'technological advances in factory monitoring systems to enable creation of new process models and support follow-on automation'. The aim of this focus area is to make use of Wireless Sensor Network (WSN) technology to improve the yield and efficiency of existing mango waste processing, thus leading to increased viability and scale of operation. This report describes the motivation for the project, the development of the WSN system, and the performance of the first 60 days of deployment.
Title: Philippines' First Wirelessly Instrumented Factory through Collaboration between CU and University of San Carlos
Author(s): Coventry University
Publication date: November 2015
Abstract:
The Philippines-UK collaborative Partnership-System for Environmental and Efficient Drying (PULP-SEED) is a British Council Newton Fund Institutional Link between Coventry University (CU), UK and the University of San Carlos (USC), Philippines. As part of this project, CU's Cogent Labs, in collaboration with the USC, have developed the Philippines' first wirelessly instrumented factory. This wireless system aims to improve the efficiency and capacity of a mango by-product processing factory.