Mahindra Logistics looks to achieve >99% accuracy in package counting and inspection across all its warehouses through AI computer vision; signs technology alliance with Jidoka Technologies

About the client – a snapshot

Mahindra Logistics Limited (MLL) a part of Mahindra Group one of the largest and most admired multinational federations of companies, is an integrated third-party logistics (3PL) service provider, specializing in supply chain management and enterprise mobility (people transport solutions). Founded more than a decade ago, MLL serves over 400+ corporate customers across various industries like Automobile, Engineering, Consumer Goods, and E-commerce. The Company pursues an “asset-light” business model, providing customized and technology-enabled solutions that span across the supply chain and people transport operations.

Client’s challenge and the business need

MLL offers warehousing services which also include inventory control and storage management.  It manages over 17 million square feet of warehousing space at multiple locations across India.  These are a mix of built-to-suit, dedicated and multi-user warehouses.  These warehouses consistently deliver the consignment within the specific timeframe.  With huge volumes of consignment movement in and out of these warehouses, it is time-consuming for the workers to keep track of the number of inbound and outbound parcels and boxes and match them against invoices/DC.  There are checks performed for damages, puffing, and leakages.  The counting and checking tasks are performed manually that also leads to many errors and inconsistencies in inventory counts and identifying of damages.

Supervisors are performing these repetitive tasks such as counting which is subject to fatigue and inaccuracies.

The Client wanted a cognitive-based cargo counting system to record case counts, damages, and invoice matching for all incoming and outgoing shipments while ensuring traceability to Invoice/DC in the WMS systems.

Scope

• Automation of the counting/inspection process.
• There was a need for a solution for inbound and outbound material or boxes which helped in achieving near-100% accuracy in package counting and inspection.
• In addition, inspect the quality of cartons for leakage, tear, etc.

Solution

Jidoka Technologies which was one of the cohorts of the CATAPULT 2.0 incubator program (a part of MLL’s ambitious tech-based initiative), was given the task of delivering the AI-based visual inspection solution to address the need.

Jidoka’s team of Engineers had discovery sessions with the client’s logistics and warehousing team to understand the requirement before developing the solution. The team had to create a system for 100% digital inspection and visibility of shipments at the time of receipt and dispatch.  This had to keep in view, count, barcode/QR code detection, SKU type detection, and box damage condition and detection, while ensuring the best fit in palletization.

Jidoka developed a Machine Vision based hardware setup.  The software was integrated with Programmable Controllers (PLCs).  These processors enabled the running of automation programs and played a key role in the automation of the warehouses.  A non-intrusive camera was set up for inspection.  AI-driven models were leveraged to mimic human counting thereby providing consistency and efficiency.  The solution was made mobile across docks. This was first piloted at the distribution warehouse in South India.

The solution evolved as the users started to see the system in action both from automation and software perspectives.

Business Benefits

• >99% accuracy was achieved in package counting and inspection. This ensured quality checks for damages on boxes detected either at the inbound or outbound stage.
• The supervisor’s productivity was increased by 50% with zero supervision on counting and quality checks, thereby providing up to 20% savings on labor costs was accomplished.
• It was non-intrusive and therefore performing at or better than human speeds
• Automated reconciliation against WMS/ERP system was established.
• The solution was mobile and deployable.
• Traceability was achieved.

About Jidoka Technologies:

Founded in 2018, by technology leaders, Sekar Udayamurthy, Dr. Krishna Iyengar, and Vinodh Venkatesan, Jidoka Technologies, the Chennai-based startup is a leader in the field of automated cognitive inspection delivering cutting-edge engineering solutions.

Jidoka stands for a principle that advocates ‘intelligent automation’ or ‘automation with a human touch’, in Japanese. Jidoka ‘s QC solution primarily consists of two parts – the first being the software platform called Kompass that connects real-time decision-making to state-of-the-art AI, to create an end-to-end system for visual defect detection. The second part is the hardware platform – which is productized into five design groups to ably support the decision-making.

It harnesses and mirrors human reasoning in defect detection delivering 98% or higher accuracy in the QC process and a significant increase in throughput.

Jidoka Tech works with manufacturers across automotive, FMCG, warehouse and logistics, pharma, general manufacturing, electronics, textiles, and printing industrial domains currently and will expand to other verticals, in the future.

Please visit, https://jidoka-tech.ai/ for more information.

______________________________________________________________________________