Midterm Study Guide 1. Articles The Southwest Secret: Why does Southwest Airlines have profit year after year with such low fares? 39th consecutive year of profit Their operations are simple: they have only one type of airline fleet, they’re the same shape, size, their mechanics are only neWe also discuss several other topics like What are exact numbers?
If you want to learn more check out biol1009
If you want to learn more check out What is any subject or part of a sample space?
If you want to learn more check out bio 315
Don't forget about the age old question of halogenation of methylbenzene
Don't forget about the age old question of How do you calculate rectilinear motion?
eded to be trained for one type of plane (ground and onboard crew). They don’t assign seats (if they swap planes and seat config is different, it doesn’t matter). They allow free checked bags (less time checking bags at the gate). They don’t have hubs, they go from pointtopoint. They don’t have international flights. Apple’s SupplyChain Secret? Hoard Lasers Wanted to have lasers but they couldn’t shine through metal. They had to make small holes to let a specific laser light to shine through. The problem was that there were too many and it was too expensive. There ops: closed ecosystem where they exert control over nearly every piece of supply chain from design to retail store Allows massive product launch They purchase all air freight in advance at Christmas to their stores are always filled Double capital expenditures on its supply chain The demand is tracked by store/hourly, teams are deployed to never run out Starbucks Complaints of not having fine coffee – mechanized process Changed the store efficiency: steam milk for individual drinks instead of pitchers, use of only one espresso machine, says might create longer lines, x2 the time Transition time: long, eventually reduces errors, but longer customer wait time Longer time to make coffee, also because beans were under the counter Changed to a consistent pace, operational changes to make coffee faster BUAD 311 Textbook (2.1, 2.2, 3.1) Presbyterian Hospitals in Philly have instantaneity of business, so use of process analysis Represented on Gantt chart Chart with all process steps and durations (activity/processing times), shows dependence between various process activities (for example, the consultation can only happen after the consultation…), used in process management Important to look at the big picture: in hospital, there are multiple patients, Gantt didn’t take into account hospital wait; equipment and people necessary are process resources, and waiting time occurs when several patients are competing for the same limited resources Process: resources that transform inputs into outputsProcess flow diagrams allow you to see the entire system of the process Flow unit: what flows (ex: in hospital, the patients) # of flow units in process: inventory or WIP Flow time: time it takes to get through the process taking into account wait Flow rate: rate at which the process delivers output Capacity: max rate with which process can generate supply Profit perspective: higher flow rate = more revenue Capacity constrained: sufficient demand that you could sell additional output Process flow diagrams are a graphical way to describe process Aggregate level: inventory plant finished goods Process boundaries = appropriate level of detail (i.e. Engineers use detailed description of steps) Process flow diagram is collection of boxes, triangles, and arrows Boxes are for activities, required for completion of flow unit, have a capacity, carried out by resources Triangles are buffers, hold inventory (does not add value), no capacity, multiple flow unit types Arrows take you through the process, multiple flow unit types possible, indicated flow of flow unit BUAD 311 Textbook (3.2, 3.3, 3.4, 3.5) Flow rate: Minimum (available input, demand, and process capacity) If demand < supply: demand – constrained If supply < demand: supply – constrained Time to produce certain amount of supply is: time to fulfill x units = x divided by flow time Utilization = flow rate divided by capacity how much of a resource is actually being used (in %) if demand < supply: process not running at full capacity, only producing demand rate if in sufficient supply of inputs, process unable to operate at capacity if process steps have limited availability, process might operate at full capacity sometimes and other times not at all (reasons why not 100% utilization) Bottleneck is resource with lowest capacity Resource with highest utilization Utilization can never exceed 100% but implied utilization can Implied utilization = demand divided by capacity captures mismatch between what could flow through the resource and what the resource can provide if exceed 100%, resource does not have the capacity to meet the demand Possible to have multiple implied utilization over 100% but there will still be one bottleneck Mysterious Persistence of CronutNY in Gilded Age: vulnerability to hype and willingness to pay obscene prices and endure penitential inconveniences Ansel bakery too small, limit of 350 cronuts a day 70% of customers are locals Process to get a cronut: two lines, 1 for cronuts, 1 for everything else Samuel: waits for you Websites satisfy Late Night Campus Snack Attacks Campus snacks: per college, limited amount of production, after businesses are closed Use of internet Taco Bell and the Golden Age of DriveThru Service Champions: specific shifts and scripts The way the script is written doesn’t put pressure on customers Orders that require ‘service champions’ because of complexity of item Drivethru: operational heart of fast food, 70% of sales, 85 cars will roll through at peak lunch time Flow very important: less orders, less production lower sales CONSISTENCY: accuracy above 90%, 164 seconds average Internet will change the notion of drivethru Little Law Is Big for Startups TTG SUCCESS: How to get organic TTG when starting a company? Traffic Traction Growth Queuing theory: Little Law Average number of items in a queuing system is equal to the average rate at which items arrive times the average time an item spends in system Queuing theory is following the arrival service departure flow L = � x W L is average number of items in queuing system (WIP) � is the average number of items arriving per unit/time (Flow rate) W is average waiting time in system for 1 item (Flow time) What matters? Not how content is distributed or price of product HOW FAST VISITORS COME AND GO? Google vs Bing? Google has highest � but low W but � is so high that L translates to 5,000/second Bing is only at 1,388/second Facebook It has a high W but wants to increase its arrival rate (�) Throughput rate only important if there is arrival 3 implications: evaluation factor: surges in traffic at much smaller time interval provide traces of higher value instead of monthly stats per second stats why and how influx of visitors in smallest time frame possible: sustaining that influx “Early traction trumps great content” normalizing metrics and looking at meaningful windows of time better than looking at longterm averages Long Line for a Shorter Wait Supermarkets – use long lines but they go faster Whole Foods started… keeping lines moving: queuing management One long line instead of small lines per register it’s more efficient and sales increase Lines can hurt retailers Once pass a certain point, people who wait more than 4.5 mins say they’ve waited 15 minutes Single line started in banks Theme Parks make sure waiting time is long on fun Instead of waiting in the hot/sun outside: air conditioned tent, with slides and climbing towers Investing in wait times to be less boring (reducing wait time whether real or perceived) Average 910 rides per day (lot of time spent in lines) Fast pass: virtual queuing Or extra cash: VIP, front of lines Interactive queue: guest satisfaction (i.e. cell phone use for games, videos, animatronic potato head) Psychology of waiting in lines Federal Express: waiting described as agonizing The first law of Science: Satisfaction = Perception – Expectation (psychological phenomenon) If P < E: dissatisfied E < P: satisfied Example: mirror installations in front of elevators to increase satisfaction Second law of service: “It’s hard to plan catchup ball” Halo Effect created by early stages of service Example: if a service experience starts badly, harder to increase satisfaction Proposition concerning the psychology of waiting: Occupied time feels shorter than unoccupied time: make times less boring, whether is it making lines into bar, making people on hold listen to stories of sports team they are calling, providing weighing machines and eye charts in hospital waiting rooms People want to get started: giving out menus in restaurant lines, gives the sense of ‘we know that you are here’, avoid the ‘fear of being forgotten’ Anxiety makes waits seem longer: choosing the wrong line… deciding whether or not to change lines, anxiety level increases. Erma Bombeck’s Law: “The other line always moves faster” Uncertain waits are longer than known, finite waits: better telling patient the doctor will be 30 minutes late than not telling him, in the latter case, patient goes through nervous anticipation Unexplained waits are longer than explained waits: if emergency is taking place in hospital and aware, you can wait with greater equanimity, in flight announcements give reason for the delay, justifiable explanations will tend to soother the waiting customer more than unjustifiable explanations Unfair waits are longer than equitable waits: the prior seating of someone who comes later than you at a restaurant causes patient customers to become furious, FIFO is appropriate rule for queue discipline, but can be broken in hospitals because of emergencies The more valuable the service, the longer the customer will wait: checkout counters in supermarkets for items of 10 or less because customers with full cart were much more inclined to wait in line, same for airlines, simple transactions vs complex = kiosk vs agent Solo waits feel longer than group waits, better to wait with group, individuals can turn to each other to express their exasperation, wonder what is happening, console each other… 2. Lecture Optimizing individual process Redesigning, coordinating systems of multiple processes There is no best way for ops, context dependent Optimizing a process to make it run smoother Disney video, improve customer satisfaction with waiting time, change the perception of it, use fast pass, they can wait for a ride while doing other rides AV testing, when you give the same group 2 tasks but with different conditions Process Analysis As unit your flowing through process as input until you come out as output Diagrams are meant to help with decision making Determine scope and level of details of flow Flow rate is (more of) an average: units are important, make sure they are all the same Capacity is the number of units that can be processed, per unit of time (rate) Difference is: the capacity is the ideally amount processed (max), but the flow rate is what is actually being processedSince flow rate is bounded by capacity, utilization rate cannot go over 100% Flow rate is on a process not a task (doesn’t vary in a process). Determined by available input, demand, and process How to improve process? add capacity to bottleneck improve balance: crosstraining best span of control stimulate demand (discounts…) Little’s Law WIP = Flow time x Flow rate Probabilities expected value variance std deviation: the spread coefficient of variation: where the values are Distributions normal/standard exponential uniform Cuttinginline Video Queuing system: arrival time service time Characteristics of queues Arrival time is not known (randomness) Service time not know either These lead to idleness of resource and wait Metrics Average waiting time, average queue length, utilization rate of server Drivers of these metrics High coefficient of variation = longer wait Utilization increase Pooling customers in one line You start waiting when capacity is too small How to compute T (time in queue), I (inter arrival time), and U (utilization)Check formula sheet. 3 key drivers: variability, utilization, risk pooling Discussion on The Goal What is the goal of your company? Raise money (depending on the company) Financial measures: net profit, ROI, cash flow Operational measures: throughput, inventory, operational expenses Production: anything that makes you closer to your goal