Dresser Rand Assembly Line MSD I Detailed Design

Dresser Rand Assembly Line MSD I Detailed Design Review Team 10458 February 18, 2010

Agenda • • • Project Current Status Risk Assessment Kitting Process and Equipment Sub Assembly Handling Software

Project Current Status • • • Crane Design Approved Kitting Cart Design Approved Bin Design to be finalized Friday, 2/19/2010 Software prototype tested once Software Specs Updated

Risk Assessment • Scope Creep (9)- OK • Kitting process • Workers Reject Equipment/Process (6)- OK • DR’s responsibility to enforce • Feedback from users • Bad Data (6)- OK

Kitting Process and Equipment Laura Mandanas and Kelly Votolato

Overview • • • Current Inventory Delivery Specifications Budget Cart Selection Approved Kitting Cart Design Proposed Kitting Cart Process Trial Cart Bin Selection Feasibility Analysis Test Plan

Current Inventory Delivery • Two rolling carts • Many small items unsorted in cardboard boxes • Inconsistent labeling (tape pull tickets/”bag and tag”, varies by operator) • Inconsistent item placement • No visual indicators of cart status Rolling carts brought from inventory to the floor.

Current Inventory Delivery Rotor subassembly on pallet. VMI. • Heavy parts and subassemblies brought to the floor on pallets • Vendor Managed Inventory (VMI) out on floor for bolts, gathered by “water spider” • Supermarket out on the floor for some common parts, gathered by operators

Key Specifications Engr. Spec. # Importance Source ES 11 2 CN 6 Speed of Kit MHD ES 12 1 CN 7 Weight lifted by worker ES 15 1 CN 10 Kitting must accommodate all parts in assembly ES 16 2 CN 12 CN 9 Height of device ft 6 <5 Requested by Matt Corman ES 17 1 Cost $ 250, 000 <250, 000 10, 000 for Kitting Process ES 19 1 CN 11 Specification (description) Safety Violations Unit of Measure Marginal Value Ideal Value ft/min 20 12 lbs 50 <30 Boolean Comments/Status 1 0

Budget • $10, 000 Projected Kitting Equipment Costs Item Cantilever Cart Bin Paint Bin Dividers Qty Unit Price Total Price 37 $0. 00 300 $11. 31 $3, 393. 00 5 $20. 00 $100. 00 1800 $2. 30 $4, 140. 00 $7, 633. 00

Cart Selection Final Kitting Cart Concept Selection Drawbridge Cart Cantilever Cart Baker's Cart Workstation Carts Criteria Importan ce Rating Score (Rx. I) Rati ng Score(Rx I) Rating Score(Rx. I) Ease of Implementation 4 1 4 5 20 2 8 1 4 Safety+Ergonomi cs 5 3 15 4 20 Footprint 1 3 3 4 4 2 2 Visibility 3 5 15 4 12 3 9 5 15 Cost 2 2 4 5 10 3 6 2 4 TOTAL 41 65 Ratings are based on a scale of 1 -5, where one is the least favorable, and five is the most favorable. 42 45

Approved Kitting Cart Design • Two rolling cantilever carts labeled by contract # – Main assembly cart – Pre-assembly cart • Removable plastic bins, labeled and color coded by station • Specific slot for each item in bin, labeled with name and picture • Flag system to indicate cart status

Proposed Kitting Process • For each contract, workers will pull parts by station and place in plastic bins – Includes VMI, which is moving back to stores • Filled plastic bins go on cantilever cart in staging area, yellow flag raised • When all parts have been pulled, green flag is raised to show that cart is complete • Operator moves cart from staging to assembly area

Trial Cart • Used during test of new smart card ticket system • Existing cantilever cart and small plastic bins from stores • Plastic bins labeled by station; color coded labels only • No dividers; items loose in bins, pull tickets placed on top Main assembly trial cart.

Bin Design • Color coded labels not visible enough; bins themselves should be color coded • Loose items in bins are not easily identified and can be damaged; need specific slots Loose items in bins.

Bin Prototype • Used pulled parts from smart card ticket test, laid out slots for each item 22 in • Substitute material (foam and 4 in cardboard); actual bins will be heavy duty plastic • Will show workers in inventory area 17 in to get feedback on basic design Shallow plastic bin with optional dividers, already available on the market. • Looking for similar items already available on the market

Feasibility Analysis

Test Plan • March (1) ▫ Usability- Pickers, Assemblers, Waterspider ▫ Audit Process Flow • April (2) ▫ ▫ Bin Durability (ES 15) Cart speed (ES 11) Max Weight Lifted by Worker (ES 12) Height of Cart (ES 16) • May (1) ▫ Carts accommodate all possible parts (ES 15)

Gantry Crane Joe Bykowicz

Overview • • Specifications Budget Concept Selection Detailed Drawing Usage Feasibility Analysis Test Plan Mechanical Simulation of Boom

Specifications Spec Number Importance Source Spec Units Achieved ES 9 1 CN 5 Load Capacity >2000 lbs Yes ES 16 2 CN 12+CN 9 Height <6 ft Yes, on all but largest units ES 17 1 Cost Total <$250, 000 Yes, 2 Units @ ~$1500 Each ES 18 1 CN 8 No attachments to foundation Boolean Yes ES 19 1 CN 11 Safety Violations Boolean Yes

Budget • $10, 000 per unit Allowance • Estimated Unit Cost= $1, 500 • Cost minimized by using existing hoists • 2 to 3 cranes needed • Estimated Total Cost= $4, 500

Concept Selection Bridge Crane Criteria Importanc e raw extende d Gantry with manual hoist raw extende d Gantry with assisted hoist raw extende d Articlated arm raw extend ed Ease of Use 8 8 64 6 48 8 64 Mobility 10 2 20 9 90 6 60 5 7 35 3 15 Safety 10 7 70 8 80 6 60 Cost 7 4 28 8 56 7 49 9 9 81 8 72 7 63 6 2 12 9 54 8 48 Quantity Needed Strength and Stability Ease of Implemenation Total 310 425 444 359

Detailed Drawing

Usage • Current Process for assembly leaves gantry crane relatively stationary • In stations 2 and 3, a rotor cart (already owned) is wheeled under the gantry crane. Rotor is lifted, and main assembly cart is moved underneath gantry. Rotor is installed. Nozzle rings and top casings are installed in similar fashion. • In this usage threat of having suspended parts swaying during movement is minimized. In case crane needs to be moved while items are suspended, crane will have eye hooks welded to the frame, operators can use supplied tow straps with rubberized hooks to secure suspended items for movement

Feasibility Analysis • Max Static Deflection= ~. 025” • Max Static Stress = ~12, 950 psi • Factor of Safety =~ 2. 75 • Based on Hand Calculations

Test Plan • March (1) ▫ Important components of test run will be ease of use and general acceptance of device by line workers. • April (2) ▫ Expected Loads will be ~400 lbs for the rotor, ~900 lbs for the top casing and ~50 lbs for the Nozzle Ring, well below Max Rated capacity. • May (1) ▫ Test any upgrades to the design based on previous feedback

Mechanical Simulation of Boom

Software Leo Gala

Overview Current State Specifications Concept Selection Detailed Design Feasibility Analysis Test Plan

Current State Automation in place. No integration that produces desired Pick Lists. Desired State Diagram.

Desired State MSD Scope

Specifications Engr. Spec. Importance # Source Specification (description) Unit of Measure Marginal Ideal Value Comments/Status ES 4 3 CN 2 Manageable file size. KB 25, 000 <2, 000 N/A ES 5 1 CN 2 Computational time Min 1 <. 5 Close to marginal. ES 6 3 CN 3 Manageable file size KB 50, 000 <25, 000 N/A <. 5 Dependant on Pick List Size. May be infeasible. 1 Achieved. ES 7 1 CN 3 Computational time ES 8 2 CN 3 Travel matrix Min Boolean 1 Scale: 1 (Important) – 3 (Convenient)

Concept Selection- Software Platform Excel & Word Database & Excel Custom Software Ease of use 5 4 4 Ease of implementation 4 4 1 Software Availability 5 3 2 Ease of Design Maintenance & Upkeep 5 3 2 3 3 1 Cost 5 4 3 27 21 13 Sum: Scale of 1 -5 with 5 being the best.

Concept Selection. Optimization Excel Solver Packages Heuristic Ease of use 4 3 3 Ease of implementation 4 2 5 Software Availability 5 2 5 Ease of Design 3 3 3 Maintenance & Upkeep 3 1 3 Optimum Solution 4 5 3 Solve Time 3 4 5 Cost 5 2 5 31 22 32 Sum: Scale of 1 -5 with 5 being the best.

Detailed Design

Feasibility Analysis Software will eliminate manual Pick List development. File manipulation should take less than a minute during run time. Straight forward navigation. Invalid Contract # will cause macros to malfunction. SD-1: Dependant on Standard Work generation. SD-1: Part-Asm. Station relationship needs to be predefined. SD-2: Optimization should only take a few minutes (pick list dependant) during run time. SD-2: Distance matrix will need to be updated as storage locations are added/modified.

Test Plan March (1) Test Sub-design 1 on several orders. (Subject to Std Work Availability). March (19) Test Pick order algorithm on sample Contract. April (2) Verify Sub-design 1 & 2 integration into DR- Intranet.

Questions?