Electric vehicles (EVs) are becoming increasingly attractive for last-mile delivery due to their low operating costs and reduced emissions. However, widespread adoption is hindered by high upfront investments, limited public charging infrastructure, and long charging durations. To address these challenges, we explore a complementary strategy that enables EV fleets to generate revenue through vehicle-to-grid (V2G) energy transactions, while improving charging accessibility through on-the-go solar energy harvesting (SEH). In this work, we formulate a vehicle routing problem for an EV fleet equipped with V2G and SEH capabilities with the objective of reducing the overall delivery costs, while accounting for vehicle load capacities and customer time window constraints. We propose a learning-to-optimize approach (LA) that scales efficiently to problem instances involving hundreds of customer locations and multiple discharge station visits for V2G operations. Using the Solomon benchmark datasets, we compare the performance of the proposed LA with a genetic algorithm (GA). Our results show that LA achieves reasonable solution quality while being 18 times faster than GA, demonstrating its effectiveness for large-scale EV fleet route optimization.

We consider the problem of scheduling resources with monetary transfers among agents in a setting where multiple outlets can dispense these resources at different rates within fixed time-slots. This problem is motivated by applications such as electric vehicle (EV) charging where energy is the resource and EVs are available within a convenient time window of its owners. The agents' valuations depend on the contiguous time slots at a given outlet that dispense the resource to them. We show that for monotone and its sub-class of dichotomous valuations, computing the social welfare-maximizing allocation is NP-hard, even if there is only one outlet. For monotone and dichotomous valuations, we provide a randomized 2-approximation mechanism that is truthful in dominant strategies and individually rational for a single outlet and a randomized O(root vertical bar S vertical bar)-approximation algorithm with the same properties for multiple outlets (S is the set of time-slots). However, for single-minded valuations, the welfare maximization problem for multiple outlets is in P. This allows us to use standard mechanisms like VCG to ensure truthfulness and individual rationality.

The shift to electric vehicle (EV) fleets has turned energy cost management into a key operational priority for logistics companies. As EVs increasingly serve both as transportation assets and mobile energy carriers, energy demand has expanded beyond (static) facility loads to include (dynamic) bidirectional charging across urban locations. This shift introduces variability in overall energy demand as EV charging requirements are tightly coupled with routing decisions that, in turn, are influenced by the cost of charging at both enterprise-owned and third-party stations. The resulting dependency between routing and energy procurement becomes even more complex when sourcing from a diverse energy portfolio that includes wholesale markets, retail contracts, and renewable sources. In this paper, we address the problem of jointly optimizing energy procurement and EV fleet routing for enterprise operations, where energy demand and source selection are cyclically dependent. To ensure scalability across diverse energy sources, service locations, and EV fleet sizes, we propose a hybrid iterative optimization framework that makes the problem tractable in real-world settings. Our method leverages the strengths of two metaheuristics to exploit the structural characteristics of the respective subproblems: (i) Genetic Algorithm (GA) for optimizing energy procurement and (ii) Coronavirus Herd Immunity Optimizer (CHIO) for optimizing energy procurement EV fleet routing. Experimental results show that the solution quality of the hybrid GA/CHIO framework is comparable to that of an exact (but computationally intractable) Mixed-Integer Nonlinear Programming (MINLP) model at small problem scales. At larger scales, it significantly outperforms a coupled GA/GA baseline both in solution quality (35 % better) and speed (55 % faster).

Unauthorized or improperly constructed speed breakers on city roads are a major contributor to slow moving traffic and road accidents in developing nations, especially in rapidly urbanizing regions. The existing practice of manual inspection by road safety personnel has not been effective due to the large scale of this problem. To this end, we present RoadHAWK: an edge-centric, automated, real-time system to detect and characterize speed breakers. We propose an efficient sensor informatics pipeline that leverages the accelerometer sensor of a vehicle mounted smartphone, and overcomes measurement and ego-noise to not only detect the features of interest but also their temporal span; both of which provide the necessary estimates to identify potential violations in the design or placement of speak breakers. On-the-road experiments show that RoadHAWK is (on an average) 75-80% accurate in estimating speaker breakers and associated violations.

Managing the bill of energy without impacting business operations is a critical requirement for large enterprises. Presently, the electricity demand in the logistic sector extends beyond conventional building loads to charging of electric vehicle (EV) fleets; which not only can be used for delivering goods, but also energy. Such an operations setup makes it difficult to ascertain the overall energy demand of the enterprise due to the variability induced by EV charging that depends on vehicle routing for multi-service delivery requirements; which in turn depends on the cost of charging EVs within and outside the enterprise. This impacts the power procurement strategy in the presence of multiple sources including electricity markets with variable prices. In this work, we consider the problem of joint optimization of power procurement and EV fleet routing for enterprises that exhibit cyclic dependency between source selection and energy demand. In order to solve this problem at real-life scales (in terms of number of energy sources; service locations; EV fleet size), we present an iterative framework using genetic algorithm (GA). We find that the solution quality of the GA framework closely matches with that of the exact intractable mixed-integer nonlinear program (MINLP) solution at smaller scales.

Rapid adoption of electric vehicles (EVs) can result in a significant new load for buildings. However, EVs can also help with building energy management because of their demand flexibility and ability to act as buffers through discharging. Specifically, joint control of HVAC and EVs can help reduce building energy costs under a time-of-day electricity pricing regime. Most existing works either do not consider discharging or the stochastic availability of EVs. We consider the problem of joint control of EVs and HVACs while respecting both thermal constraints of HVAC and state of charge (SoC) constraints of EV users. We complement existing works by treating EVs as buffers with random availability. We propose evhac, a multi-agent reinforcement learning framework for EV-HVAC joint control that scales seamlessly with increasing EVs. We evaluate our approach in a simulated environment calibrated with real-world building data and EV charging demand. As baselines for our approach, we use default PID control; and a model-predictive control (MPC) approach with and without using EVs as buffers. Our experiments show that 1) our approach scales significantly better (86x) than MPC in decision time, while being slightly worse (6%) in energy costs; 2) there exists a sweet spot for the cost savings potential as a function of EV and HVAC load profiles; and 3) MPC performance is very sensitive to the prediction horizon.

Demand for public charging is growing with the increasing adoption of electric vehicles (EV). Motivated by these developments, the parking industry is fast transitioning from pure play parking to parking with charging; but are facing difficulties to scale charging operations due to high infrastructure costs and grid-level capacity constraints. These challenges can be potentially alleviated by organizing the charging network into clusters, each consisting of a collection of EV chargers. To prevent imbalances between the charging demand across network clusters and available power supply, while being able to service different type of customers with priority for prime users; the power flow in the charging network needs to be managed intelligently. To achieve this, we consider the problem of optimal power allocation to not only satisfy these multiple system objectives, but also scale efficiently to large network sizes involving hundreds of clusters and charging points. We present a solution approach based on mixed-integer linear program (MILP) to overcome the above challenges. Using simulations, we show that MILP performs better than a greedy approach by allocating 12% more power to prime customers while respecting various system constraints; with solution speed ranging from a few seconds (on small networks) to less than a minute (on large networks).

Sustainability benefits and reduced operational costs are making electric vehicles (EV) the preferred choice for last-mile deliveries. The competitive pricing of wholesale electricity markets and distributed energy resource capability of EV fleets (in addition) provide a revenue channel through energy arbitrage. To capture this value, captive bidirectional EV chargers need to be managed intelligently to handle electricity price variations and the energy demand of the EV fleet. EVs need to be charged sufficiently for their complete delivery runs. Therefore, (dis)charging decisions need to intelligently exploit the limited flexibility between vehicle trips while respecting energy market commitments, and limited number of chargers that can only either charge or discharge at any given instant. We complement existing works with an energy model that uses electricity bought from the day-ahead market for charging the fleet, and uses the intra-day market for arbitrage. We overcome the challenges of problem scale and its varying dimensionality by using (i) a graph representation based learning agent (LA3_D) with two-stage encoding for day-ahead charge planning; and (ii) a priority order based greedy heuristic (GH_I) for intra-day arbitrage planning. Because the agent learns the planning policy of mapping EVs to charging operations over several problem instances, it is able to solve a given instance with limited sub-optimality when put to test at different levels of scale. Evaluation studies using representative datasets show that the proposed (LA3_D + GH_I) model for charging with arbitrage reduces costs by 17% (for mid-size fleets) and 23% (for large fleets) compared to charging without arbitrage. Results from medium-to-large problem instances without arbitrage show that the average charging cost of LA3_D is 20% better than other learning baselines; while being reasonably close (≈ 12%) to the optimal for smaller problem instances.

The use of electric vehicles (EV) in the last mile is appealing from both sustainability and operational cost perspectives. In addition to the inherent cost efficiency of EVs, selling energy back to the grid during peak grid demand, is a potential source of additional revenue to a fleet operator. To achieve this, EVs have to be at specific locations (discharge points) during specific points in time (peak period), even while meeting their core purpose of delivering goods to customers. In this work, we consider the problem of EV routing with constraints on loading capacity; time window; vehicle-to-grid energy supply (CEVRPTW-D); which not only sat-isfy multiple system objectives, but also scale efficiently to large problem sizes involving hundreds of customers and discharge stations. We present QuikRouteFinder that uses reinforcement learning (RL) for EV routing to overcome these challenges. Using Solomon datasets, results from RL are compared against exact formulations based on mixed-integer linear program (MILP) and genetic algorithm (GA) metaheuristics. On an average, the results show that RL is 24 times faster than MILP and GA, while being close in quality (within 20%) to the optimal.

Electric vehicle (EV) fleets are well suited for last-mile deliveries both from sustainability and operational cost perspectives. To ensure economic parity with non-EV options, even captive chargers for EV fleets need to be managed intelligently. Specifically, the EVs needs to be adequately charged for their entire delivery runs while handling reduced time flexibility between runs; limited number of chargers; and deviations from the planned schedule. Existing works either solve smaller instances of this problem optimally, or larger instances with significant sub-optimality. In addition, they typically consider either day-ahead or real-time planning in isolation. We complement existing works with a hybrid approach that first identifies a day-ahead plan for assigning EVs to chargers; and then uses online replanning to handle any deviations in real-time. For the day-ahead planning, we use a learning agent (LA) that learns to assign EVs to chargers over several problem instances. Because the agent solves a given instance during its testing phase, it achieves scale in problem size with limited sub-optimality. For the online replanning, we use a greedy heuristic that dynamically refines the day-ahead plan to handle delays in EV arrivals. We evaluate our approach using representative datasets. As baselines for the LA, we use an exact mixed-integer linear program (MILP) (greedy heuristic) for small (large) problem instances. As baselines for the replanning, we use no-planning and no-replanning. Our experiments show that LA performs better (8.5-14%) than greedy heuristic in large problem instances, while being reasonably close (< 22%) to the optimal in smaller instances. For online replanning, our approach performs about 7-20% better than no-planning and no-replanning for a range of delay profiles.

Deployment of the Internet of Things (IoT) in smart buildings has received considerable interest from both the academic community and commercial sectors. Unfortunately, the widespread adoption of current smart building solutions is inhibited by the high costs associated with installation and maintenance. Moreover, different types of IoT devices from different manufacturers typically form distinct networks and data silos. There is a need to use a common backbone network that facilitates interoperability and seamless data exchange in a uniform way. In this paper, we present EMIoT, a novel solution for smart buildings that breaks these barriers by leveraging existing emergency lighting systems. In EMIoT, we embed a wireless LoRa module in each emergency light to turn them into wireless routers. EMIoT has been deployed in more than 50 buildings of different types in Sydney, Australia and has been successfully running over two years. We present the design and implementation of EMIoT in this paper. Moreover, we use the deployment in a residential building as a use case to show the performance of EMIoT in real-world environments and share lessons learned. Finally, we discuss the advantages and disadvantages of EMIoT. This paper provides practical insights for IoT deployment in smart buildings for practitioners and solution providers.

This is our second edition of the Internet of Things Series of IEEE Communications Magazine. Recently, Mischa Dohler has taken over from Luca Rose as Lead Editor of the IoT series. In the meantime, Prasant's tenure has come to an end. The community and the magazine thank both for their invaluable services over the past years. After a careful selection process, we have appointed Taras to the team, who has ample experience in emerging technologies - welcome to the team, Taras.

Cyber-physical systems (CPS) are systems where a decision making (cyber/control) component is tightly integrated with a physical system (with sensing/actuation) to enable real-time monitoring and control. Recently, there has been significant research effort in viewing and optimizing physical infrastructure in built environments as CPS, even if the control action is not in real-time. Some examples of infrastructure CPS include electrical power grids; water distribution networks; transportation and logistics networks; heating, ventilation, and air conditioning (HVAC) in buildings; etc. Complexity arises in infrastructure CPS from the large scale of operations; heterogeneity of system components; dynamic and uncertain operating conditions; and goal-driven decision making and control with time-bounded task completion guarantees. For control optimization, an infrastructure CPS is typically viewed as a system of semi-autonomous sub-systems with a network of sensors and uses distributed control optimization to achieve system-wide objectives that are typically measured and quantified by better, cheaper, or faster system performance. In this article, we first illustrate the scope for control optimization in common infrastructure CPS. Next, we present a brief overview of current optimization techniques. Finally, we share our research position with a description of specific optimization approaches and their challenges for infrastructure CPS of the future.

Multiunit residential building (MURB) residents are an upcoming segment of electric vehicle (EV) owners and potential buyers (around 42% in Europe). Garage-orphaned MURB residents have to mostly rely on public chargers, which currently handle only 5% of the EV charging needs. With EVs becoming more mainstream, public chargers will not be able to match the operations scale without additional deployments. This will not only lead to a demand–supply mismatch in the short term but also impact the growth of EV adoption in the long term. For managing the demand–supply mismatch, dynamic pricing is a widely used control tool, but it is often difficult to make informed pricing decisions when 1) there is variability (both) in demand and supply, 2) users’ spatiotemporal behavior and price elasticity are unknown, and 3) charging preconditions (such as the state of charge) are not freely available. In this article, we present SurCharge, which uses reinforcement learning (RL) to overcome these challenges in dynamic pricing for EV charging. Our approach is evaluated on real-world traffic patterns for Luxembourg by augmenting the Luxembourg Simulation of Urban Mobility traffic scenario simulator with EV charging demand models. The results show that the proposed RL-based SurCharge system delivers a 10%–24% higher revenue margin than other competitive dynamic pricing baselines, without making unrealistic assumptions of prior models and data.

The articles in this special section focus on the application of Internet of Things to sensor networks. The number of Internet of Things (IoT) technology applications has been growing steadily. IoT devices can perform basic functionalities for sensor data collection and machine control. Furthermore, advancements in chip design have enabled sophisticated system-on-chip applications over small-sized IoT devices that perform complex data processing at the network edge. Unlike human-operated devices, such as smart phones, the operation of IoT devices is often expected to be autonomous. However, the massive numbers of IoT nodes in large-scale communication networks pose new challenges for IoT system design and management. Presents selected novel applications, highlights essential security and privacy aspects, and evaluates the performance of the emerging cellular technology that supports IoT connectivity.

In this paper, we propose a spectrum sharing protocol for cognitive radio networks with an energy-constraint primary transmitter (PT), which performs energy harvesting from the received radio frequency (RF) signals broadcasting by two secondary users (SUs). After the amount of harvested energy at the PT is sufficient for data transmission, a SU will cooperatively forward PT's signal using amplify-and-forward (AF) scheme along with transmitting its own signal to increase the spectrum efficiency of the system. To be specific, the data transmission block can be divided into two phases. During the first phase, the PT will optimally select a relay from nodes S1 and S2 in order to further improve the overall throughput, then the PT transmits primary signal to the selected SU and primary receiver (PR), respectively, while the non-selected SU transmits signal to secondary receiver (SR). In the second phase, the selected SU transmits signals of both the primary and secondary systems simultaneously. A discrete Markov chain is used to simulate the charging and discharging processes of the PT's battery. In order to avoid mutual interference between the primary and secondary systems, a fraction of spectrum is used to transmit primary data and the residual spectrum is served for data transmission of the secondary system. An appropriate bandpass filter (BF) is deployed at the receive front-end of each receiver to extract the desired signals from assigned bandwidth. Based on this, the exact expressions of the outage probabilities for both the primary and secondary systems are derived. Moreover, the optimal bandwidth allocation coefficient is determined by utilizing convex optimization to maximize the achievable rate of the secondary system while guaranteeing the achievable rate of the secondary system in information transmission mode. Numerical results show that the proposed spectrum sharing protocol outperforms to other scheme and non-cooperative scenario in terms of average spectrum efficiency.

In many metropolitan cities, multi-unit residential buildings (MURB) are becoming more common than single-family independent homes due to lack of urban space. MURB residents (around 42% in Europe) are potential adopters of electric vehicles (EV), but lack a private garage for EV charging. They need to exclusively rely on public charging, which currently serves only 5% of EVs. As EVs become more prevalent, the lack of extensive public charging can create a short-term demand-supply mismatch in specific city neighbourhoods, as well as preclude long-term growth in EV adoption. We believe that uberization of private garage chargers that are typically under-utilized during day-time can alleviate this problem. In this work, we examine how a charging service provider can match public charging demand with private suppliers while using a demand-response based pricing model. We base our study on real-world traffic patterns for the city of Luxembourg by augmenting the Luxembourg SUMO traffic scenario (LuST) simulator. Specifically, an EV's charging demand is modeled by a state machine with charge/discharge dynamics based on Tesla Model-S. Our preliminary results suggest that the proposed uberization strategy has the potential to gracefully handle demand spikes with higher revenue yield for a charging service provider, even while handling different categories of service users.

In recent times, unmanned aerial vehicle (UAV) also known as drone has been extensively used to study the scope of such technology in a range of applications including emergency communication coverage, agriculture, search and rescue, temporary wireless communication coverage, packet delivery, smart city deployments, defense, etc. Acknowledging the features offered by UAVs in different application domains, here we describe a vision for leveraging UAVs with wireless sensor networks to enhance the ability of mining operations and to provide a safe work environment to the miners in both routine and emergency scenarios. The main contributions of this chapter are threefold. First, it critically presents the potential applications of UAVs in mine and their use cases. Second, it provides and discusses the basic wireless networking architecture for such high-stress work environment. Furthermore, different design challenges are also discussed in detail. Third, considering a mine disaster scenario, we propose a UAV based multi-hop emergency communication system. Different performance metrics such as packet error rate, delay, and number of retransmissions per packet have been used to study the proposed emergency communication framework. The proposed system will assist the miners and rescue team members in case of an emergency and improves the quality of experience to the end users.