To Be or Not To Be Automated – Lets Find The Answer!
Published on : Monday 02-11-2020
The decision to automate your processes depends on multiple factors, says Paras Arora.
With all the taglines reading, ‘robots are here’, they often fail to recognise the unique challenges and complexities occurring in individual industries and geo-economics of a country like India. Robotic Process Automation or simply automation, as referred to within this article, can potentially bring reduction in operational costs, increased reliability and prepare you for an unprecedented scale. It is although a long arduous, and often one-way journey rather than a switch one can turn On or Off. Even the decision to disrupt the current manual processes to an automated one, brings headaches to many supply chain managers, and rightly so.
So, how to decide whether one should start investing in robotics and automation? While I am a celebrator of technology and robotics, my experience at Flipkart has taught me the importance of stepping back and acknowledging the human versatility. Following is a study of comparing the fit of humans and automation and understanding the transition from the former to the latter, especially with the scope of hard automation and not just software automation.
Scale
The most important factor towards automation is the scale you are operating on. Few questions you should ask yourself and the planning team:
1. What is the frequency of transactions or actions to be taken on each of the given item type?
Demand or production planning for any logistic or manufacturing asset, along with a Time Motion Study, will help you arrive at this number. If, for example, you need to perform a given task on the assembly line only once in 5 min, a complex and automated machinery specialised for the task is overkill, but where any operation needs to be performed tens or hundreds of times on a given car within a min, for example tack welding on a car chassis, it pushes the utilisation of the robotic system close to 100%, making it much more reasonable.
Similarly, a local FMCG brand, operating on low margins, catering to a few hundred households at best or serving an exclusive clientele, for example personalised toiletries, are better served by a few expert hands. On the opposite spectrum, the likes of HUL or Leher, etc., would want to automate as much as possible to produce an output in thousands per minute.
2. How many differentiated SKUs, in terms of shapes and sizes are there?
When it comes to material handling, one size does not fit all. There is nothing as dexterous as a human hand. Imagine an item with any shape or size (within certain limits, of course) humans will be able to handle it, but the same cannot be said for robots. E-commerce logistics controls some of the largest and busiest warehouses today. The gamut of items handled by e-commerce companies prevents them from complete automation. Instead, it is incremental and only certain processes are automated. On the other hand, a large medicine or a nut bolt distributor can easily automate all the storage and warehousing needs, even the quality control through vision-based systems.
3. Is there a segregation of the different kinds of SKUs? Can they be bucketed together?
While the items of varied shapes and sizes cannot be handled at piece level by automated systems, binning – as in logistics – or kitting – as in assembly lines – them together allows partial automation. Suddenly the world opens up to solutions like goods-to-man, sorters, palletizers, etc.
4. Do you have the space to scale up?
With shrinking availability of commercial space and added requirement of social distancing it is not always practical to expand horizontally. Either grow vertically or improve space utilisation in given facilities. Beyond 12 metres for manual operations, regulatory requirements play a major role requiring fire hazard management and multiple exits. This leads to further reduction in overall space utilisation. Automated storage and retrieval systems, sortation solutions and conveyorisation, all help in achieving both the factors
Dynamicity
Humans, who have the capability to learn something new every day, do multiple differentiated tasks, move around and change roles. Robots on the other hand believe in the philosophy of doing one thing and doing that well.
Automated systems often require drastic changes, including layout changes and physical hardware installation. Even with the planned growth, it is not always possible to simply add ‘more’ automated systems to increase capacity. During sale period when peak demand is in multiples of that during usual business, quickly varying capacity is often not possible with automated systems.
In real world, fluctuations in demand and supply occur and new bottlenecks appear. People can be retrained and asked to handle different roles and perform complex tasks. An automated system on the other hand would have to be powered down or procured, as the requirement may be.
Results from automation are often a discrete function, rather than a continuous one. Thus, a good balance has be struck, between automated and manual process and even areas have to be dedicatedly allotted for future expansion. This balance can look extremely different for different industries, depending on the slope of growth or peaks.
Task Complexity
Want to read barcodes on items? Conveyor based fixed scanners reign. Want to read labels affixed on items, at approximately well-defined locations, vision systems can do > 1/sec. For placing the items in the right orientation but, both depend on humans. To fill a crate with 300 ml glass bottles, no manual intervention in required; but to do the same with vegetables, unless you like your vegetables perforated or squished, let a person do the job.
As a thumb rule, higher the complexity, higher is the manpower requirement. So, breakdown your process into smallest actions, see which ones are nothing but repeated motions on repeated items, and you might just have found the ideal candidate for automation.
Costs
A ubiquitous concern across industries, across countries even, before making that decision on whether to automate or not, is cost. Robotic installations are often capital expenditure (CapEx) heavy, with payback period generally ranging from 3 to 15 years, depending on the technology and manpower cost, along with its span. Often ignored savings of an automated/robotic system comes in the form of real estate and manual overheads saving. Au contraire, often the installation cost, pilot cost of robotic systems and margins for delay do not feature in the initial business case which leads to frustration later. One such example is civil work that you come to know once the contract is awarded. Moreover, you should not compare automation with as is process but the most optimised version of manual processes that you can obtain, given the identical constraints. For example, if you plan to automate picking for AA category (highest speed) movers, efficiency of manual processes only for this limited category should be evaluated instead of an overall average.
Technology readiness
Remember, automation is a journey and not simply a destination. It begins with understanding how well established your internal systems are, and what is needed to start making the change. Taking an example from logistics, having mobile robots for moving items around, you need to have a well-designed Warehouse Management System, the civil work done as per the requirements of the robots, redesigning of processes, retraining of people, all will be required to welcome automation. Depending on technology and the process, and your company’s current state, this preparedness can take from couple of weeks to a few months. In general, it is always a good practice to keep upgrading your current processes and training to the best available benchmarks, so when the need for automation arises, the transition is smooth.
Social impact
Social responsibility is not a question of ‘when’ but ‘if’. Automation, without doubt, leads to job losses in the short run, requiring people to upskill or shift industries. Companies and industries geared towards providing employment or helping the blue-collared work force, will have to answer this question as part of policy even before evaluating any automation solution.
A contrary version of it is applicable in hazardous industries where certain jobs are riddled with life threatening situations. In such a case, automated systems are a recommendation, even if it is not the most cost effective.
The illustration is a simple diagram to help in making the decision on whether to automate or not. If the answer to any of the question is no, you should hold the decision and work on its readiness. To summarise, the decision to automate your processes depends on multiple factors like your status quo, where you want to be, and of course, your value system. While Robotic Process Automation is indeed the future of multiple manufacturing units and supply chains, it does not come without its complications. People, process, facility and bank balance are all impacted by it. Appreciating its complexity and building the right decision tree is the first step towards a successful turnkey project. How to go from 0 to 1, and handling the transition is a topic for another time.
Paras Arora has a passion for Robotics, Automation and IoT, and has almost 10 years of experience in supply chain and manufacturing domains. He has a data and process first approach and is versed with lean principles, Go to Market strategy and Product Market Fit. Although he is more focussed towards logistics industry, his varied experience has allowed him to plan and execute projects in B2C and B2B industries, across private and public sector, in various companies and start-ups, like NTPC, Flipkart, Delhivery. He has completed his Masters in Advanced Robotics from European scholarship program and Bachelors in Mechanical Engineering from BITS Pilani.