What you need to prepare for the year 2020.
IoT could have as great an impact on transportation as the invention of shipping containers and the advent of overnight delivery services by DHL, UPS, and FEDEX.
Seismic shifts in technology like this drives some businesses out of business and creates opportunities for existing or new ones. Amazon.com for example could not have grown into such a global phenomenom without Fedex. And when the first two international airlines, PanAm and TWA, could not adapt to deregulation and the resulting drop in air fares, they went bankrupt.
Market forces like these force transportation companies to continually look for new ways to lower costs in order to stay competitive. Inefficient operators are taken over by their competitors. Globalization exposes carriers and fleet operators to competition from low cost rivals in other countries.
Now with profit margins driven almost as low as they can go, the companies that survive today will be those that use IoT and big data to optimize operations.
Let’s take a look at what specific items transportation companies can quantify in order to run as lean a business as possible. We focus on truck fleet operations, but these ideas apply equally to passenger and cargo airlines, taxis, buses, ocean shipping, and local delivery vans.
Trucks are the perfect IoT business use case. Because there is the need to monitor vehicles on the road from a central location. The tools to do that are the IoT cloud, big data databases, analytics, the ERP system, onboard sensors, and wireless communications.
There are many metrics to measure on a moving truck. Intel says that the average heavy truck generates 10MB per kilometer of data.
IoT serves several goals. First is the need to do preventive maintenance to keep vehicles in service. There is also the need to change the delivery route in real time in order to meet shifting customer requirements, road conditions, etc. Plus there is the goal of operating the vehicles safely, adhere to environmental emissions regulations, and conserve fuel.
Manufacturers already ship many of their machines and devices with IoT onboard. Rolls Royce and other engine manufacturers monitor engines for their customers. Goodyear and other tire makers put RFID chips in their tires.
As for the fleet operators themselves, they can locate GPS, heat, humidity, vibration, and other sensors on brakes, engines, refrigerated trailers, lights, etc. Data gathered there is uploaded to the IoT cloud using a cellular connection.
How does the fleet operator use this data?
The biggest change in fleet operations is the advent of big data and analytics software in the cloud. This lets trucking companies process streaming and unstructured data and run algorithms across that. The output of this analysis creates work orders in the ERP system, sends the driver notifications, and lets dispatchers and managers capture what is happening in real time on dashboards.
What does that mean?
Consider the situation 10 years ago. Trucking companies might have used satellites to upload data or downloaded metrics manually from their vehicles when they returned to the trucking hub. The data was loaded into data warehouses so that BI (business intelligence) could be run over that. All of that was inefficient because it took lots of time to convert data to the data warehouse format.
Now free opensource databases like Spark, Storm, Kafka, and Hadoop let companies process this data in real time and require no, or little, data conversion. And the company no longer needs to hire an army of statisticians. That is because statistical models are built into programming languages that let regular programmers and ordinary end users gain insight from data. And for companies without programmers at all, they can upload that data to cloud companies to do that for them.
To give an example, consider the classic transportation problem taught in business schools. The goal of that is to use linear programming to set the route for the whole fleet, meeting customer deliveries, yet drive the shortest distance. That is how routes were set before big data and IoT. Now on-board metrics and the big data cloud can have trucks change course in route as conditions change in real time. For example one supplier’s warehouse has run out of inventory, another truck has cracked a brake drum, and one driver has driven more hours than the law allows.
Another goal is to keep the vehicle in operations. One problems trucks have is the need to change brake pads on 18 wheels. In the absence of IoT, companies might do this on a fixed schedule. Thus they are changing brakes too often and wasting money. With IoT a company can monitor brake rotor temperature and kilometers driven and use predictive models to accurately show when it it time to change brakes. Scania trucking company describes the problem here. And here is Spark ML code to show how to run a predictive model against brake data coming from a streaming IoT application and expose that as a web service.
IoT too can show which drivers are operating their vehicles in an unsafe manner or in a way that is more likely to damage the engine and drivetrain and waste fuel.
There is another changing factor that is going to force transportation companies to adopt IoT. While no one is talking about driverless heavy trucks yet, the fleet of driverless passenger vehicles is growing. Those cars will be equipped to talk to each other, so it is only natural that they talk to trucks. And the highways themselves will become smart with sensors embedded in the roads to report on road conditions and more. Cities too already stream data on traffic conditions as do crowdsourced applications, like Waze.
So trucking companies and transportation companies of all type need to prepare for IoT now. Their competitors are doing that, and changing technology means it will be forced upon them. Manufacturers are already building IoT into their machines, so the trucking companies can tap into that data as their build their transportation models. And the growth of IoT and big data means that standards are emerging so that the cost and complexity of doing all of that is going down. So it is a win-win situation for the companies that move to adopt that and a threat to those who do not.