ICT DEPLOYEMENT STRATEGY IN AQUITAINE WSC: THE EXPLOTIC PROJECT BREAKTHROUGH Adrien Arraiolos a, Morgan Vuillermoz b, Maryse Bigot b a Research Engineer Institut Technologique FCBA Allée de Boutaut BP 227, F - 33028 Bordeaux cedex Email: [email protected] b Research Engineer Institut Technologique FCBA 10 avenue Saint Mandé, F – 75 012 Paris ABSTRACT French forest-based companies are currently facing an important challenge, the need to increase wood mobilization in order to answer the growing energy and construction markets. The collective target is to harvest an additional volume of 20 millions cubic meters per year by 2020. The challenge is even greater in Aquitaine region. Economic crisis, two destructive storms and recurring insects’ attacks will disturb the forest industry for the next ten years. Fighting against these hazards requires innovative business strategies and solutions to improve the day to day efficiency of logging and supply operations. Better cooperation supported by tools enabling anticipation, agility and real time operations, is the key to achieve this challenge while ensuring added-value creation. ICT deployment is believed to pave this road towards competitiveness as novel technologies will provide operational knowledge-based indicators for forest-based companies. A coherent and pragmatic extended enterprise development strategy is necessary for successful ICT integration. This extended enterprise concept will allow a demand-driven business model and maximize the value to be generated. A collective R&D strategy was designed by FCBA with its professional partners to meet these challenges. Each project part of this dynamic relies on end-users, their vision and their needs. A major milestone was reached with the ExploTIC project successful end. This project delivered an open IT architecture. Based on StanforD files, this architecture allows a more accurate harvesting site monitoring and short time control. Keywords: ICT deployment, StanforD data exchange, Adoption, Wood supply chain, Harvester data. INTRODUCTION French National Forest Policy calls for an increase of the wood harvest to face the growing demand from the construction market and the development of the use of wood for energy. In the Aquitaine region, in South-West of France, this challenge is amplified by the critical consequences of the economic crisis and two large windstorms which hit the region during the last decade (MARTIN 1999 and KLAUS 2009). On a regular basis, 8 Mm3 of Maritime Pine would be harvested in the region as a fair balance to the annual biological growth (Colin et al. 34th Council on Forest Engineering, June 12-15, 2011, Quebec City (Quebec) 1 2010) and as to fit wood demand requested by the regional forest-based industry (GIP ECOFOR 2010). But because of the last windthrown and consecutive forest decay due to insects’ invasions, it was determined that the annual wood harvest would have to come down to 5Mm3 for at least the five up-coming years (Cavaignac 2010). Stocks available on local wood storage areas and softwood importation from other regions will help reduce this deficit. However, the risks are high that a supply crisis will quickly weaken the Industry and deeply destabilize the existing WSC. FCBA and its professional partners believe that a way out of this situation would be to: - Change the way businesses operate into more flexible and adaptive processes; - Enable better data flow within and between companies to support the upper-mentioned organisation shifts. Novel process organizations and technologies are key factors of these improvements. Hence, a RTD strategy was elaborated and implemented to propose new methods and solutions to industrial users. This paper will present the chosen methodology and highlight the results already gained from EXPLOTIC, one of the RTD project composing the strategy. CONTEXT Wood supply organization in Aquitaine Forests in “Landes de Gascogne” are mainly covered by Maritime Pine (85% of standing wood volumes, IFN 2010), on flat terrain and sandy soils. Since the 1999 storm, mechanization was considerably developed. 85% of the trees are now harvested by CTL machinery (Emeyriat et al. 2009). RoleCompany FoFowwoorrnneeeessrrttss mmaannaaggeerrssFFoorreesstt ccoommppaanniieesspprrooccuurreemmeennttWWoooodd CCoonnttrraaccttoorrssLLooggggiinngg ccoommppaanniieessTTrraannssppoorrtt MMiillll FFoorreesstt oowwnneerr FFoorreesstt mmaannaaggeerr LLooggggiinngg Main role mmaannaaggeemmeenntt LLooggggiinngg Complementaryrole ooppeerraattiioonnss TTrraannssppoorrtt sseerrvviiccee WWoooodd ttrraannssffoorrmmaattiioonn Figure 1 : Roles played by different types of company in the WSC Rows define the different roles that wood supply chain companies (columns) can assume in Aquitaine. Like in many other areas in France or other countries, logging operations are mainly done by contractors since wood procurement companies now tend to prefer an outsourcing system. 34th Council on Forest Engineering, June 12-15, 2011, Quebec City (Quebec) 2 Many actors are involved in the WSC (see Figure 1) from the forest to the mill gate and one company can often play several roles. This results in a complex but flexible organisation of the supply system. Challenges Both post-storm crises demonstrated the industry’s fragility facing uncommon events. Several weaknesses were pointed out: - a low knowledge on forest stands characteristics, - rudimentary processes follow-up methods, - impossibility to trace logs efficiently, - Inefficient supply chain logistics management. Poor data management often appeared to be causing these deficiencies which result in loss or degradation of logs, process waste, and delay in the harvest, the transport or more generally speaking the whole WSC management. Before Klaus (2009), the Industry remained competitive despite these problems thanks to the advantage created by an easy access to abundant raw materials. But now that the resource has been deeply reduced, it has become urgent to developed new strategies to harvest and use wood more efficiently. This evolution requires first an increase of day to day efficiency in the logging operation management that will finally lead to a global yield optimization from standing trees purchase to mills process. In such perspective, access to information is critical and data management must be enhanced using ICT. In practise, a long term step by step strategy based on targeted ICT integration is required to implement this new wood supply chain management method. This strategy needs to combine (Ginet et al. 2011): - Automation of standardized data transfer, - A better technical and financial monitoring of harvesting blocks, stock piles, machines, transportation services and operators. - Short time control, - Optimization of wood resource allocation by transforming wood quality characteristics expected by the process in trees description parameters (Moreau 2010), - Decision support and predictive tools to locate stands, harvest trees and deliver logs with a better match with the industries specifications (Vuillermoz & Arraiolos 2010). METHOD TO SUPPORT ICT DEPLOYMENT IN AQUITAINE Creation and implementation of the RTD strategy A collective R&D strategy was designed by FCBA with its professional partners to find solutions to the up-mentioned challenges. Each project part of this dynamic relies on end- users, their vision and their needs. Starting from a need analysis, each project goes through a technology benchmarking, a development phase followed by integration on test sites. Following this method, it becomes possible to develop prototype and test them in real conditions. Test runs results provide input for process redesign of the prototype as well as clear specifications of ICT tools needed to 34th Council on Forest Engineering, June 12-15, 2011, Quebec City (Quebec) 3 achieve the targeted redesign. With a valid proof of concept, transfer can be organised so that the novel system or solution can be offered to the industry. In Aquitaine, this integration started in 1999 and is still running. ExploTIC as a project is part of this global approach (Figure 2) and Figure 9 in APPENDIX also shows where the project stands in the overall IT deployment. 1999 2015 Wood Logisticsin Aquitaine Indisputable Key eMobois ETF Monitoring (FCBA, 1999) (UE project, 2007-2010) (FCBA, 2010-2014) (Defined) DEFOR Logistics EExxppllooTTIICC LNPKey (FCBA, 2005-2008) (FCBA, 2008-2011) (FCBA, 2011-2013) BLOGFOR FlexWood Logiscom (FCBA, 2005-2009) (FP7 project, (CAFSA, 2011-2013) 2009-2012) PPrroocceessss XX XX XX XX XX XX XX OOppttiimmiizzaattiioonn SSttaannddaarrdd ddaattaa XX XX XX XX XX XX XX XX eexxcchhaannggee TTrraacceeaabbiilliittyy XX XX XX Aquitaine ICT R&D projectstimelineand issues.A long term step by step strategy based on targeted ICT R&D projects is required to implement a innovative wood mobilization system. R&D projects deal with three types of issues : process optimization, standard data exchange and traceability. Each project has at least one direct application in Aquitaine. Figure 2 : Aquitaine ICT R&D projects timeline and issues ExploTIC project genesis EXPLOTIC started from the diagnostic that IT functionalities within the harvesting machines were completely under-used by both the contractor and the supply companies though these tools are often key facilitators towards: - Reliable data exchange between the field and the office - Better machine utilisation rate EXPLOTIC mission was to provide tools and solutions for operational production data transfer between harvesters, forwarders, contractors and wood supplying companies in order to better handle organizational issues, e.g.: - Increasing data exchange quality and reliability between field and office with any machine working in the “Landes de Gascogne”. - Decreasing wood supply companies’ employees road travel of 50%, - Improving machines utilization rate. A collective project from the start, ExploTIC associated representatives from the Aquitaine Industry (Smurfit Kappa Comptoir du Pin, Forestière de Gascogne, CAFSA, Castagnet- Duméou), machines manufacturers (Ponsse and John Deere) public authorities (Agriculture Ministry and Aquitaine regional council) and FCBA (acting as coordinator). 34th Council on Forest Engineering, June 12-15, 2011, Quebec City (Quebec) 4 Method applied during the EXPLOTIC project From the very beginning, the project scope was clearly bounded: - The solution shall be build on a maximum number of existing technologies - It should be interoperable with every machines working in the region. The project began with an end-users needs analysis. Ten contractors, four supply companies and two forest machines manufacturers were interviewed. Directly involving contractors in an IT project was a first for those professionals who are often seen as too small business to participate in such dynamics. In addition to this analysis, a technology benchmarking was run on existing harvester-embedded software and foreign practices. A trip in Finland was organised with a group of users to see on in the field how StanforD standard is implemented. The next step consisted in designing the IT Architecture that would enable standforD- compatible harvester-data exchange while overcoming the lack of inter-operability from the IT component installed in the machines at the time. Exchange sequences were built for logging sites organisation, implementation and monitoring. A working group including end- users was involved in this architecture and software specification work. The latter was mostly dedicated to choosing StanforD variables and transforming it in a relational database. This task was based on UML sequence, use case and class diagrams. Once specified, development started in an iterative way under close watch from the end-users group. Easy-to-use guidelines were produced to explain how to run the IT system on a regular basis. The ExploTIC IT architecture was then tested and validated on a test site of more than 20 machines, 5 contractors, and 4 supply companies ( Figure 3). Implementation involved wood procurement companies and logging machines owners distributed in four sub-groups – each with its own characteristics – in order to cover most “Landes de Gascognes” cases ( Figure 3). Supply Machines Implementation characteristics Machines characteristics Companies owners 1 CAFSA AFM Pereira The contractor works for several sub- Network connected Ponsse SARL Palomo contractors and to wants to use machines StanforD data to its own Non-computerized machines 2 Forestière de SARL Valeiro A lot of contractors working mainly Network connected Ponsse and Gascogne SARL Eurobois with forestière de Gascogne John Deere machines SARL Labat Non-computerized machines 3 SKCDP SKCDP A mix of wood procurement company’ Network connected John Deere SARL Deyres machines and contractor machines. A machines (D an E series) compatibility with Smurfit group data Non computerized excavators security strategy is required with felling head. 4 Société Société The wood procurement company only Not connected John deere Castagnet- Castagnet- works with its own machines. machines Duméou Duméou Non-computerized machines Figure 3 : Test sites organization Tests sites requested regular intervention and monitoring in order to: 1. Analyse the existing sequences and the means used to exchange data, 34th Council on Forest Engineering, June 12-15, 2011, Quebec City (Quebec) 5 2. Update the existing system by connecting available IT, integrating the missing ones and redesigning data model 3. Train the operators on the easy-to-use guidelines, 4. Control all the data exchanged over the IT Architecture and also received by FCBA office, 5. Assist operators to improve their skills in the system use, 6. Gather enhancements requests. 7. Specify and follow-up debug developments requested to the software services company. Field tests were also a good opportunity to run a cost/benefits analysis and identify the benefits the system would bring to the logging process. Potential gains can trigger user interest. It was therefore a prerequisite to the design of the broader deployment strategy that was finally presented and validated by both professional and instutional representatives. RESULTS Lesson learnt from the Need Analysis First analysis revealed the incompatibility between harvesters’ software from different manufacturers. At that time (2008), Timbermatic 300, Timberoffice 4.0 and Opti 4G 4.6 were the last versions of the software available. They were able to write StanforD data, but they did not meet the standard format. Hence writing and sending a file from one brand’s software made it impossible to read in another. Some variables were mandatory for the exchange but were defined as optional in the software interface. This standardization failure was obviously calling for the specification and development of an interoperable solution. The needs analysis also showed that a special attention should be paid to the project acceptability by contractors. Their main reluctance came from the fear that time monitoring done by the harvester computer would be transferred to sub-contractors. Such data exchange would reduce contractors’ autonomy and negotiation capacity on contracting rates. An other initial barrier was the logging contractors’ lack of awareness of what could be gained from by better management of production-data. The needs analysis was concluded by a redesign of logging operations monitoring and control procedures based on the use of StanforD data exchange. ExploTIC IT architecture StanforD data included in apt, prd, prl and drf files were fully parsed and analysed to understand which variables would fit the businesses needs. The needed information were sorted into groups (Figure 4) and structured in a relational database model. Each table attribute from this model is linked to StanforD variables ( Figure 5). 34th Council on Forest Engineering, June 12-15, 2011, Quebec City (Quebec) 6 Data group Role Details Structuring data Allow data aggregation by creating an unique Wood products referential for each company Species Product groups Machines Persons and companies Site management data Data required to plan and prepare a site Harvesting sites data Bucking Instruction Production order Field data Data sent back from machines Harvester production Forwarder production Time monitoring data Figure 4 : ExploTIC data groups Analyzed data were sorted in order to clearly identify which ones met the industry needs. Time monitoring data Attribute name StanforD variable ID Explotic internal variable activityDate Var53_t1 activityDuration Var316_t1:t23 or var317_t5 or var319_t5 Depends on activityTypeID CreationDate Explotic internal variable activityTypeID Var316_t3 or var317_t3 or var319_t3 Depends on activityTypeID OperatorID Var212_t1 MachineID Var3_t1 SiteID Var21_t1 Figure 5 : An exemple of data modelization and StanforD linkage Each ExploTIC database model table attributes is defined and linked, when it is required, to StanforD variables. These information ensure the compatibility with the standard. This data analysis was approved by the end-users group and was then used as a basis for the definition and development of the system architecture. The decision was taken to specify and develop a StanforD compatible software. Data transfers between existing machines’ IT systems and ExploTIC software were designed (Figure 6) and development was assigned to SDIGIT (a forestry software services company). The delivered ExploTIC software enables: - Field data creation and exportation in a StanforD format for non-computerized machines’ operators, - StanforD bucking instruction (APT files), harvester (PRD files) and forwarder (PRL files) production and monitoring data (DRF files) importation and aggregation inside a unique database for machines and logging managers. - Connection to other databases through an opened SQL engine. 34th Council on Forest Engineering, June 12-15, 2011, Quebec City (Quebec) 7 EXCHANGE SYSTEM LOGGING COMPANY OFFICE MACHINE (WEBMAILS / USB) Production data aggregation software Osonf tbwoaarred ONRDSATFNRCEAETFI PMBrooundciutkocinrtiigno gnIn d*s.atprturadc ,*t .i*od.nrpfrl ONRDSATF RNRASETFProduction *.prd, *.prl ONRDSATF NRCEAETFI Business DBAUTSAINBEASSSE *.apt features Production *.prd, *.prl Bucking Instruction Monitoring data *.drf *.apt STANFORD INTERFACE Production data aggregation software Business features BUSINESS DATABASE CONTRACTOR OFFICE Figure 6 : ExploTIC ICT architecture The architecture model shows the different data exchange possibility and the means used between the involved actors. The major innovation is the connector role of this application. The database is opened (firebird base) and the SQL engine easily allows new additional modules for data treatment and connection to any information system. Training tasks were anticipated by writing easy-to-use guidelines, based on screenshots and illustrated procedures. A pilot to demonstrate EXPLOTIC efficiency These experiments showed that the system could cover most of the possible cases: from companies equipped with different machines (Ponsse, John Deere…) to non-computerized ones (Figure 7). 34th Council on Forest Engineering, June 12-15, 2011, Quebec City (Quebec) 8 Figure 7 : ExploTIC system adaptation to any cases The four cases treated covers the different machines configurations met on the “Landes de Gascogne” Forest. These tests were extended and successful in 2011 on new machines series (Opti 4G 4.7 and Timbermatic H09). During these tests, bugs and features requests concerning ExploTIC software but also John Deere and Ponsse solutions were identified and relayed to the development teams involved. Finally, the conclusion of this pilot is an operational system and directly measured benefits. Measured benefits and costs Once connected to data treatment system the ExploTIC architecture can provide the following improvements: - Accurate harvesting sites profitability analysis, - Day to day production quality control with time expenditure reduction, - Detailed operators time sheet justification, - Contractors services rates negotiation tools, - Logging operations customization depending on sites characteristics, - Machines preventive maintenance alerts, - Price matrix optimization, - Roadside stock piles volumes anticipation. ExploTIC leads to a better match between harvest and mills needs and gives an increased operational efficiency in logging operations management, both for contractors and wood- supplying companies. Regarding the adoption challenge, operators were less reluctant to use data transfer after a few weeks of use. Such is showed by the following testimony by Loïc, an operator from Castagnet-Duméou Company: «First, I was thinking that ExploTIC was only used by my boss to control my work. After two weeks, I realized that it reduces the time spent each week to count my work time and production volume and to write it on my diary – especially as the company clerks must type this data again at the office. But, it was even more useful the day I had to prove that my productivity was decreased because the former week, I had been ordered to change three times of site and I had spent more time in roads travel. Finally, this system is as useful to me as it is to my boss.» The deployment cost was also estimated. The identified expenditure items are operator time consumption, machine immobilization, consultant services, training sessions and software investments. An average calculation shows that it will cost 9 600 € to a contractor owners of three machines (Figure 8). 34th Council on Forest Engineering, June 12-15, 2011, Quebec City (Quebec) 9 Machines expenditure items Operator time 1.5 days in training 2 000 € consumption and 0.5 day of machine immobilization for IT Machine immobilization deployment Consultant services 1.5 days per machine 1 200 € Training session 1.5 days 1 200 € Software investments ExploTIC : for free 0 € TOTAL 4 400 € Office expenditure items Executive time 1 day in training 2 000 € consumption 1 day for data redesigning Consultant services 3 days per company 2 400 € Training session 1 days 800 € Software investments ExploTIC : for free 0 € Integration to the existing information 0 € for excel connection system until 40 000 € for a complete integration into an ERP TOTAL From 5 200 € to 45 200 € Figure 8: Detailled expenditure items Deployment cost is calculated by estimating the time spent and IT investment cost. This data were gathered during on-field tests. In France, professional training is well subsidized and such collective project can be partly founded by public authorities. This cost will be reduced to an estimated cost of 2 500 € for each logging contractor. Thus, this cheap deployment investment encourages the ExploTIC system adoption by small businesses. EXPLOTIC AND AQUITAINE ICT DEPLOYMENT OUTLOOK Next steps towards a real-scale ExploTIC deployment in the overall Aquitaine region The next step consists in deploying this IT system on 200 forest machines, 80 contractors and 4 main wood supplying companies. This deployment will be completed in 3 phases: - Industrialization of the ExploTIC software to make it more robust and consolidate its functional scope, - Training engineering in collaboration with a professional training centre and involving directly forest machines operators and logging operations supervisors, - Train users and deploy the system within 2 years. Training of users, software installation, support to company data redesign and software updates will be included in a global ExploTIC services package. Each deployment on a given machine or in a given office will require the global package to be successful. ExploTIC project position in the overall Aquitaine IT deployment The ExploTIC is a breakthrough in the whole round wood mobilization system. Future RTD projects will be able to build on this IT architecture when even they require digitalized data transfer between offices and machines. 34th Council on Forest Engineering, June 12-15, 2011, Quebec City (Quebec) 10